bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

APOE4 is Associated with Differential Regional Vulnerability to Bioenergetic

Deficits in Aged APOE Mice

Tal Nuriel1,2,*, Delfina Larrea1,3, David N. Guilfoyle4, Leila Pirhaji5, Kathleen Shannon6, Hirra Arain1,2,

Archana Ashok1,2, Marta Pera1,3, Qiuying Chen7, Allissa A. Dillman8, Helen Y. Figueroa1,2, Mark R.

Cookson8, Steven S. Gross7, Ernest Fraenkel5,9, Karen E. Duff1,2,10, †, and Estela Area-Gomez1,3, †

1Taub Institute for Research on Alzheimer's Disease and the Aging Brain, Columbia University, 630 West 168th Street,

New York, NY 10032, USA.

2Department of Pathology and Cell Biology, Columbia University, 630 West 168th Street, New York, NY 10032,

USA.

3Department of Neurology, Columbia University, 630 West 168th Street, New York, NY 10032, USA.

4Center for Biomedical Imaging and Neuromodulation, Nathan S. Kline Institute, Orangeburg, NY 10962, USA.

5Department of Biological Engineering, Massachusetts Institute of Technology, 350 Brookline Street, Cambridge,

MA 02139, USA.

6Animal Facility, Nathan S. Kline Institute, Orangeburg, NY 10962, USA.

7Department of Pharmacology, Weill Cornell Medical College, 1300 York Avenue, New York, NY 10065, USA.

8Cell Biology and Gene Expression Section, Laboratory of Neurogenetics, National Institute on Aging, National

Institutes of Health, Bethesda, Maryland, 20892

9Broad Institute, 415 Main Street, Cambridge, MA 02142, Cambridge, Massachusetts, 02139, USA.

10Division of Integrative Neuroscience in the Department of Psychiatry, New York State Psychiatric Institute, 630

West 168th Street, New York, NY 10032, USA.

* Corresponding author: Tal Nuriel, email: [email protected]

†These two authors contributed equally to this work and are co-senior authors.

bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

ABSTRACT

The ε4 allele of apolipoprotein E (APOE) is the dominant genetic risk factor for late-onset Alzheimer’s

disease (AD). However, the reason for the association between APOE4 and AD remains unclear. While

much of the research has focused on the ability of the apoE4 protein to increase the aggregation and decrease

the clearance of A, there is also an abundance of data showing that APOE4 negatively impacts many

additional processes in the brain, including bioenergetics. In order to gain a more comprehensive

understanding of the APOE4’s role in AD pathogenesis, we performed a multi-omic analysis of APOE4 vs.

APOE3 expression in the entorhinal cortex (EC) and primary visual cortex (PVC) of aged APOE mice.

These studies revealed region-specific alterations in several bioenergetic pathways, including oxidative

phosphorylation (OxPhos), the TCA-cycle and fatty acid metabolism. Follow-up analysis utilizing the

Seahorse platform revealed decreased mitochondrial respiration in the hippocampus and cortex of aged

APOE4 vs. APOE3 mice, but not in the EC of these mice. Additional studies, as well as the original multi-

omic data suggest that bioenergetic pathways in the EC of aged APOE mice may be differentially regulated

by APOE4 expression. Given the importance of the EC as one of the first regions to be affected by AD

pathology in humans, this differential bioenergetic regulation observed in the EC vs. other brain regions of

aged APOE4 mice may play an important role in the pathogenesis of AD, particularly among APOE4

carriers.

INTRODUCTION

Possession of the ε4 allele of apolipoprotein E (APOE) is the major genetic risk factor for late-onset

Alzheimer’s disease (AD). In normal physiology, the apoE protein plays a vital role in the transport of

cholesterol and other lipids through the bloodstream, as well as within the brain (1-3). Although the three

common isoforms of apoE (E2, E3 and E4) differ from each other by only two amino acids—apoE2

(Cys112, Cys158), apoE3 (Cys112, Arg158), apoE4 (Arg112, Arg158)—this small change in amino acid bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

sequence has a large effect on protein structure, causing the apoE4 protein to possess a more globular

structure due to increased interaction between its N- and C-terminal domains. This difference in structure,

in turn, affects the type of lipids that each isoform binds, as well as the affinity of each isoform for the

numerous receptors that mediate the uptake of apoE and its cargo into cells (see reviews by (1, 4)).

Importantly, these isoform differences also have a major impact on the pathogenesis of late-onset

AD. Although the APOE2, APOE3 and APOE4 alleles are present at a relative frequency of about 8%, 77%

and 15% in the normal population (5, 6), the APOE4 allele is present at a relative frequency of 36-64% in

AD patients (7-10), with individuals who possess one or two APOE4 alleles having a 3- or 12-fold increased

risk of developing AD, respectively (11). Although a number of mechanisms have been proposed to help

explain this APOE4-associated susceptibility to AD, the precise cause remains a source of debate. The

prominent hypothesis is that this increase in AD risk is due to the ability of apoE4 to increase the

aggregation and decrease the clearance of A (12-18). However, APOE4 expression has also been shown

to have deleterious effects on numerous A-independent pathways, including lipid metabolism, tau

pathology, bioenergetics, neuronal development, synaptic plasticity, the neuro-vasculature, and neuro-

inflammation [see reviews by (19) and (20)], any number of which could play a pivotal role in the

pathogenesis of AD among APOE4 carriers.

In terms of APOE4’s effects on bioenergetics, a number of pivotal reports have demonstrated that

APOE4 expression leads to widespread dysregulation of the brain’s bioenergetic capacity. For example,

early reports by Eric Reiman and colleagues demonstrated that both young and old APOE4 carriers display

decreased glucose utilization, as measured by fluorodeoxyglucose positron emission tomography (FDG-

PET), in similar brain regions to that seen with AD patients (21-23). Additional reports detail the wide-

range of bioenergetic insults that APOE4 expression can cause in the brain, including impaired insulin

signaling (24, 25), reduced cerebral blood volume and cognitive function in response to a high fat diet

(HFD) (26, 27), altered genetic expression of glucose-regulating enzymes and transporters (28, 29), and the

generation of a toxic C-terminal fragment of apoE4 that can directly target electron transport chain (ETC) bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

complexes in the mitochondria (30-32).

In order to study the diverse effects that APOE4 expression has in the brain, we performed an initial

multi-omic study of differential APOE isoform expression in an AD-vulnerable vs. an AD-resistant brain

region from aged APOE targeted replacement mice, which express human APOE in place of their mouse

Apoe gene and do not develop overt AD pathology. The EC was chosen as our AD-vulnerable brain region

because it is one of the first brain regions to develop AD-related tau pathology, in humans (33, 34), and

also because the morphology of the EC has been shown to be specifically affected by APOE4 gene

expression. (35, 36). The PVC was chosen as our AD-resistant region because it is relatively spared in AD

(33, 37, 38). As described below, these studies revealed APOE4-specific alterations in the levels of

numerous genes and small-molecules related to energy metabolism, which were explored further using the

Seahorse platform. In agreement with previous data, this analysis revealed APOE4-associated deficits in

mitochondrial respiration in several brain regions. However, contrary to what we observed in other brain

regions, the EC displayed increases in oxidative metabolism in the context of APOE4 expression,

suggesting that, under metabolic stress, brain region-specific counterbalancing mechanisms are in play in

the EC of aged APOE4 mice. Additional studies and further mining of the multi-omic data was then used

to elucidate this differential bioenergetic regulation in the EC of aged APOE4 mice, which may play a

causative role in the pathogenesis of AD and may point to novel therapeutic interventions for preventing or

slowing the disease.

RESULTS

Multi-omic analysis reveals differential expression of energy-related genes and metabolites in the EC

of aged APOE4 mice

In order to investigate the effects of APOE4 expression in an untargeted manner, we performed a multi-

omic analysis (transcriptomics and metabolomics) on RNA and small-molecules extracted from an AD-

vulnerable brain region (the EC) vs. an AD-resistant brain region (the PVC) of 14-15 month-old APOE bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

targeted replacement mice. For the transcriptomic analysis, RNA was Trizol-extracted from the EC and

PVC of aged APOE mice (10 APOE3/3 and 19 APOE3/4 males). For the metabolomics analysis, small-

molecules were extracted from the EC and PVC of aged APOE mice (8 APOE3/3, 9 APOE3/4 and 7

APOE4/4 males) using a methyl tert-butyl ether (MTBE)/methanol extraction protocol modified from

previous reports (39, 40). The RNA and small-molecules were then analyzed using RNA-sequencing and

untargeted metabolomics, respectively.

The data generated from these transcriptomics and metabolomics analyses (summarized in Tables

1 and 2 and Supplementary Tables S1 and S2) have recently been published as parts of larger studies on

APOE4’s impact on endosomal-lysosomal dysregulation (41) and neuronal hyperactivity (42), respectively.

However, as is often the case with omics and other systems biology approaches, the large amounts of data

generated from these analyses inform us about changes in multiple biological pathways that cannot all be

investigated in one study. In the case of the transcriptomics analysis, in addition to the dysregulation of

endosomal-lysosomal genes that were revealed in our pathway analysis (and which we investigated further

for our previously published study), another major KEGG pathway that was observed to be significantly

enriched for differentially expressed genes was “Oxidative Phosphorylation” (Fig. 1A). This was driven by

a large number of differentially expressed genes encoding for subunits of ETC complexes I-V, with each

of these genes being upregulated in the EC of aged APOE3/4 mice, as compared to aged APOE3/3 mice

(Fig. 1B). This increased transcription of ETC genes may also be related to the increases in energy-related

metabolites that we observed in the EC of aged APOE4/4 mice (which we reported as part of our neuronal

hyperactivity study). These metabolites include several TCA cycle metabolites (malate, citrate and

isocitrate), as well as fructose-6-phosphate, carnitine, and ATP, each of which had higher levels in the EC

of aged APOE4/4 vs. APOE3/3 mice (Fig. 2A). Furthermore, the metabolomics data indicates a higher

ATP:ADP ratio in the EC of aged APOE4/4 vs. APOE3/3 mice (Fig. 2C), which suggests an increase in the

rate of oxidative metabolism in this region. Thus, in addition to the APOE4-associated differences in

endosomal-lysosomal regulation and neuronal activity that these original studies highlighted, the data from bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

both the transcriptomics and metabolomics studies also pointed to specific differences in several

bioenergetic pathways.

Intriguingly, our metabolomics data also revealed an APOE4-associated decrease in the levels of

numerous free fatty acids in aged APOE4/4 vs. aged APOE3/3 mice (Fig. 2B), which, unlike most of the

previously mentioned energy-related metabolites, were similarly differentially altered in both the EC and

the PVC. Although it is disputed whether or not free fatty acids can be utilized in the brain as an alternative

energy source via -oxidation (43-45), there is compelling evidence to suggest that it can (46-48). Thus, it

is tempting to speculate that the decreased fatty acid levels observed in both the EC and PVC may be a

result of increased -oxidation in both of these brain regions, which may also be related to the observation

that carnitine, a metabolite that is necessary for importing fatty acids into the mitochondria for energy

utilization by the carnitine-palmitoyl transferase system, was the only energy-related metabolite in Fig. 2A

whose levels were shown to be higher in both the EC and the PVC of the aged APOE4/4 mice.

Seahorse analysis reveals decreased mitochondrial respiration in the PVC, hippocampus and cortex

of aged APOE4 vs. APOE3 mice, but not in the EC.

In order to validate and expand upon the bioenergetic differences that we observed in our multi-omic

analysis, we conducted a series of Seahorse experiments on mitochondria isolated from the EC and other

brain regions of aged APOE4/4 vs. APOE3/3 mice. Utilizing the Seahorse XF24 platform (see methods

section), we first measured the oxygen consumption rate (OCR; a measure of OxPhos efficiency) of

mitochondria from the EC and PVC of 15-month-old APOE mice (4 APOE3/3 and 4 APOE4/4 males;

samples pooled within each genotype and region). As shown in Supplementary Fig. S1, we observed a

significant reduction in both the Complex I-mediated and Complex II-mediated OCR of mitochondria from

the PVC of the APOE4/4 vs. APOE3/3 mice, but no difference in the OCR between genotypes in the

mitochondria isolated from the EC region. In order to investigate whether this decreased OCR in the PVC

was unique to the PVC or whether it represented a more global decrease in mitochondrial respiration within bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

the aged APOE4 brain, we performed an additional experiment measuring the OCR of mitochondria

isolated from the EC, hippocampus (Hip) and cortex (Ctx) of 20-month-old APOE mice (2 APOE3/3 and 2

APOE4/4 males; samples pooled within each genotype and region). As shown in Fig. 3, we observed a

significant reduction in the OCR of mitochondria from the Hip and the Ctx of APOE4/4 vs. APOE3/3 mice,

with the most abundant differences observed in the Complex I-mediated OCR. However, in mitochondria

from the EC of APOE4/4 vs. APOE3/3 mice, we did not observe any significant differences in Complex I-

mediated basal mitochondrial respiration or in complex II-mediated OCR, and we observed significant

increases in State 3 respiration in Complex I-mediated OCR in mitochondria from the EC (Fig. 3B).

Furthermore, using the data generated from this Seahorse analysis, we determined the respiratory control

ratio (RCR) (state3u/state4o), which is a measure of the coupling efficiency between OxPhos and ATP

production (49, 50) (Fig. 3C). This analysis indicates that EC mitochondria from APOE4 mice are highly

coupled, resulting in increased production of ATP per unit of oxygen.

These observations suggest that, while APOE4 expression leads to an overall reduction in

mitochondrial function in the brains of aged APOE4 mice, a result consistent with the APOE4-associated

bioenergetic deficits observed using alternative approaches (24-29), mitochondria in the EC seem to

undergo differential bioenergetic regulation, characterized by increased coupling efficiency. Importantly,

these region-specific differences in the effects of differential APOE isoform expression on mitochondrial

respiration did not appear to be the result of changes in mitochondrial mass, as the levels of Tom20 protein,

as well as the levels of PGC1α RNA and the ratio of Mitochondrial:Nuclear DNA, were unchanged between

APOE genotypes in the EC, Hip, and Ctx of 21-month-old APOE mice (Supplementary Fig. S2A, C & D).

Differential APOE isoform expression also did not result in changes in the levels of specific ETC complex

proteins in the EC, Hip or Ctx (Supplementary Fig. S2B).

Computational analysis of untargeted metabolomics data reveals differential expression of ketones

in the EC of aged APOE4 mice bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

In order to investigate the source of these differential bioenergetic effects in the EC, we conducted a

series of additional analyses on the EC of aged APOE4/4 vs, APOE3/3 mice. First, we utilized the

PIUMet (Prize-collecting Steiner forest algorithm for Integrative Analysis of Untargeted Metabolomics)

method (51) to predict identities of the untargeted metabolites that we uncovered in the EC of aged

APOE4/4 vs. APOE3/3 during our original metabolomics analysis. PIUMet uses network optimization to

analyze the data in the context of a vast database of protein-protein and protein-metabolite interactions,

revealing both known and uncharacterized pathways that contain the putative metabolites and other

molecules. PIUMet also allows for the incorporation of additional datasets, allowing us to utilize our

targeted metabolomics results in the analysis.

We used PIUMet to analyze the 304 untargeted metabolite features whose levels were differentially

altered between APOE4/4and APOE3/3 in the EC of aged APOE mice (corrected p-value < 0.05). Among

these features, 124 had matches in PIUMet’s underlying database of metabolites (Supplementary Table S5).

PIUMet was able to infer the identity of 32 metabolite features, and revealed a network of protein-protein

and protein-metabolite interactions associated with their dysregulation. The resulting network is enriched

in 18 putative GO biological processes (Table 3) and 124 total metabolites (Supplementary Fig. S3). Among

these, the most intriguing GO biological processes were fatty acid metabolic process, IMP metabolic

process, ketone catabolic process, steroid metabolic process, and vitamin metabolic process. The most

interesting putative identifications were acetone, adenylosuccinate, alpha-tocotrienol, and oleamide. Alpha-

tocotrienol is a vitamin E metabolite, which bolsters the observation in the targeted metabolomics result

that vitamins and vitamin-derivatives (including tocopherol, another vitamin E metabolite) were found to

be increased in the EC and PVC of aged APOE4/4 mice in our targeted metabolomics analysis (Tables 1

and 2). Similarly, the changes in the levels of adenylosuccinate support the idea that the purine nucleotide

cycle is affected in APOE4/4 EC, as this metabolite is downstream of IMP and upstream of

succinoadenosine, both of which were identified in our targeted analysis (Table 1). Importantly, the

identification of acetone and of the ketone catabolic process GO term may also represent a substantial

finding. Ketone bodies (acetone, acetoacetone and -hydroxybutyrate) are primarily generated in the liver bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

from fatty acid breakdown via -oxidation, and are used as a supplemental energy source when glucose

levels are low. These ketone bodies readily cross the blood brain barrier and can act as a potent fuel for the

brain (52-54). Alternatively, there is evidence that astrocytes also have the ability to produce ketones (55-

57), both through the breakdown of fatty acids, as well as through the breakdown of ketogenic amino acids,

most notably leucine (58, 59).

Proton nuclear magnetic resonance spectroscopy analysis reveals differential expression of

phosphocreatine, glutamate and GABA in the EC of anesthetized aged APOE4 mice.

1H NMR spectroscopy provides in vivo measurements of brain metabolites and has been used in many

studies of neurodegenerative disorders in both humans and rodents. This method allows for the

simultaneous acquisition of several highly abundant metabolites in the brain, including neurotransmitters

such as glutamate and gamma-aminobutyric acid (GABA), as well as other biological metabolites such as

creatine, glutathione, and N-acetylaspartate (NAA). It also allows for the accurate quantification of

individual metabolite concentrations. Since metabolomics and 1H NMR spectroscopy have been described

as complimentary techniques, with advantages and disadvantages to both [see review by (60)], we chose to

perform 1H NMR spectroscopy on the EC of 18-19 month-old APOE mice (7 APOE3/3 and 7 APOE4/4

males) in order to gain additional information about the metabolic changes occurring in the EC of aged

APOE mice. While the EC is located in a very NMR unfriendly brain area because of its close proximity to

sinus spaces, we were able to achieve full width half maximums of 8 – 15 Hz using a customized second

and third order shim coil. Fig. 4A shows the position of the volume of interest (VOI) that was used to

measure metabolite levels in the EC of these aged APOE mice. A representative spectrum from this region

is shown in Fig. 4B, along with the spectral position of some of the major metabolites. To our knowledge,

this is the first in vivo 1H NMR study of the EC in mice.

As shown in Fig. 4C and Table 4, this 1H NMR study revealed significant differences in the levels

of multiple metabolites in the EC of aged APOE4/4 vs. APOE3/3 mice, including decreased levels of

creatine (Cre), glutamate (Glu), GABA, and taurine (Tau) in the EC of aged APOE4/4 mice, as compared bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

to that of aged APOE3/3 mice. Perhaps the most interesting observation, however, was the large increase

in the levels of phosphocreatine (PCr) and the ratio of PCr:Cr observed in the APOE4 EC. Creatine is

mainly phosphorylated in mitochondria, where creatine kinase uses the newly synthesized ATP to produce

PCr and ADP (61). This PCr is then shuttled to the cytosol where it acts as a rapidly mobilized reserve of

phosphate groups that can be used to replenish ATP levels, whereas the increased levels of ADP in the

mitochondria will stimulate oxygen uptake and oxidative metabolism, especially during enhanced neuronal

activity (62). Thus, the observation that PCr levels are increased in aged APOE4/4 mice supports and

expands upon the bioenergetics observations from our transcriptomics, metabolomics and Seahorse

analyses. In addition to these changes in PCr levels, the observation that glutamate and GABA are both

decreased in the EC of APOE4/4 mice may also be related to these bioenergetic differences between APOE

genotypes and brain regions. Although glutamate and GABA are mostly known for their roles as

neurotransmitters in excitatory and inhibitory neurons, respectively, they can also be utilized as energy

metabolites (63-66).

Taken together, our data suggest that, contrary to other brain region such as the cortex and

hippocampus, the EC of aged APOE4/4 vs. APOE3/3 mice exhibit an increased rate of oxidative

metabolism and ATP turnover, which may play a causative role in the pathogenesis of AD.

DISCUSSION

Possession of the APOE4 allele greatly increases an individual’s risk of developing AD. Although

numerous theories have been proposed, the precise cause of this association remains unknown. In order to

further investigate the effects of APOE4 gene expression in AD-relevant brain regions, we initially chose

to perform a multi-omic analysis on RNA and small-molecules extracted from an AD-vulnerable brain

region (the EC) vs. an AD-resistant brain region (the PVC) of aged APOE targeted replacement mice. In

addition to our recently published reports describing APOE4’s impact on endosomal-lysosomal

dysregulation (41) and neuronal hyperactivity (42), these studies also revealed APOE4-associated bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

alterations in the levels of numerous genes and small-molecules related to energy metabolism (Fig. 1 and

2). Follow-up experiments utilizing the Seahorse platform revealed APOE4-associated deficits in

mitochondrial respiration in several brain regions, including the PVC, the Hip, and the Ctx as a whole,

which is consistent with previous reports showing that APOE4 expression has numerous deleterious effects

on the brain’s bioenergetic capabilities (24-29). Intriguingly, however, this APOE4-associated decrease in

mitochondrial respiration was not observed in the EC, with Complex-1 mediated respiration actually

increasing in the EC of 20-month-old APOE4 mice. (Fig. 3). Additional studies, including proton nuclear

magnetic resonance (1H NMR) spectroscopy and a bioinformatics analysis of our untargeted metabolomics

data provided further support and elucidation of this region-specific bioenergetic regulation by APOE4,

including the observations of increased PCr levels and decreased glutamate and GABA levels (Fig. 4), as

well as an upregulation of the ketone pathway (Table 3), in the EC of aged APOE4 mice.

Bioenergetic regulation in the brain is a complex process involving multiple interconnected

pathways and cell-types, which makes it difficult to decipher the mechanism or impact of these region-

specific bioenergetic effects of APOE4 expression. It is important to note that the decline in mitochondrial

functionality that we observed in the cortex and hippocampus, while significant, is likely below the

pathological threshold (67). Mitochondria are part of a multifaceted network of metabolic pathways that

are modulated by the cell to compensate for any reductions in respiratory activity and ATP production (68,

69). Moreover, mitochondria have a “reserve respiratory capacity” that allows them to increase their

oxidative capacity and ATP production depending on the cell’s demand (70). Furthermore, mitochondrial

adaptability is tissue and cell-type specific, depending on the different compensatory mechanisms and the

availability of substrates (67, 71)

The most intriguing finding from our study, however, is that, within the brain, mitochondria appear

to have regional differences in their response to metabolic stressors, at least in this case of APOE4

expression in aged APOE mice. In both the EC and the PVC, for example, it appears that there are attempts

to use alternative fuel sources to compensate for APOE4’s effects on mitochondrial function. This is best

represented by the decreased fatty acid levels and increased carnitine levels in the metabolomics data, bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

perhaps due to increased efforts to supply the TCA cycle with additional acetyl-CoA molecules via -

oxidation of fatty acids. However, the data from our 1H NMR spectroscopy study, as well as further mining

of our transcriptomics and metabolomics data, suggest that the mitochondria in the EC of aged APOE4

mice have additional mechanisms that allow them to counteract the APOE4-associated reductions in

mitochondrial respiration. Specifically, we hypothesize that mitochondria in the EC of aged APOE4 mice

are distinctively able to increase their respiratory capacity by increasing the coupling efficiency between

oxygen consumption and ATP production.

This hypothesis is supported by our Seahorse data, where mitochondria in the EC of aged APOE4

mice show increased ADP-stimulated oxygen consumption and increased maximal respiration rate, without

changes in proton leak (Fig. 3 and Supplemental Fig. S1) or mitochondrial mass (Supplementary Fig. S2).

In addition, we also observed that the RCR is increased in the EC of aged APOE4/4 vs. APOE3/3 mice

(Fig. 3C and Supplementary Fig. S1C), further confirming that ADP-driven mitochondrial respiration in

the aged APOE4 EC is better coupled to ATP production than in the Hip and the Ctx. Interestingly, the

reductions in mitochondrial respiration observed in the Hip and Ctx of aged APOE4/4 vs. APOE3/3 mice

by Seahorse analysis appears to be more robust in complex I-mediated respiration compared to complex II-

mediated respiration, whereas in the EC, complex I-mediated respiration in aged APOE4 begins to outpace

complex I-mediated respiration in APOE3 mice with increasing age (Fig. 3 and Supplemental Fig. S1).

Since complex I-mediated mitochondrial respiration is driven by pyruvate, this suggests that the APOE4-

associated bioenergetic deficits observed in the hippocampus and cortex of aged APOE4 mice are primarily

related to decreased aerobic respiration that occurs downstream of glycolysis, whereas in the EC we observe

an amplification of this process.

While the reason for the increased efficiency in OxPhos in the EC of aged APOE4 mice requires

further investigation, we note that the expression of some mitochondrial transporter genes is specifically

upregulated in mitochondria from this brain region, as detected in the transcriptomics data (Supplementary

tables S1 and S2). In order to ensure a sufficient rate of solute flux into mitochondria to fuel metabolic

pathways, numerous mitochondrial transporters link biochemical pathways in the cytosol and mitochondria bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

(72). Regulation of these transporters is an essential part of the overall regulation of cellular metabolism,

and their expression is highly variable depending on the tissues and the metabolic conditions. Intriguingly,

the expression of the adenine nucleotide translocase (ANT) genes, ANT1 (Slc24a4) and ANT2 (Slc24a5),

were significantly increased in the EC of aged APOE3/4 vs. APOE3/3 mice. ANT participates in ATP-for-

ADP exchange through the inner mitochondrial membrane, which supplies the cytoplasm with newly

synthesized ATP from OxPhos. Interestingly, ANT can also interact with creatine kinase (CK), gaining

preferential access to ATP, which is needed to synthesize PCr in cells with high energetic needs (73).

Similarly, the expression of the genes for the voltage dependent-anion channel (VDAC1; Vdac1), one of

the main gatekeepers for the regulation of the crosstalk between mitochondria and cytosol (72), is also

increased in the EC of aged APOE3/4 mice, as is the gene for the mitochondrial phosphate transport protein

(PTP; Slc25a3), which transports inorganic phosphate (Pi) to the mitochondrial matrix in order to supply

the Pi required for ADP phosphorylation into ATP (72).

Another important mitochondrial transporter that ensures the balanced cooperation between

OxPhos and glycolysis (and to a lesser extent the pentose phosphate pathway) in order to maintain ATP

levels is the malate-aspartate shuttle. Briefly, in order for the two NADH molecules generated during

glycolysis to be translocated into mitochondria, malate must be produced in the cytosol via conversion of

oxaloacetate (OAA) by malate dehydrogenase (MDH1), for which NADH is a necessary cofactor, and then

malate is converted back to OAA by MDH1 inside mitochondria (converting NAD+ back to NADH in the

process). Since this newly formed OAA cannot be transported back to the cytosol, OAA must first be

converted to aspartate by glutamate oxaloacetate transaminase 1 (GOT1; Slc25a22). But this requires the

donation of an amino radical from glutamate. Thus, glutamate is an essential component for maximizing

the number of ATPs produced in glycolysis per molecule of glucose metabolized. Interestingly, our

transcriptomic data reveals that the expression of both Slc25a22 (GOT1) and Mdh1 are upregulated in the

EC of aged APOE3/4 vs. APOE3/3, suggesting that activity of the malate-aspartate shuttle is increased in

the EC of these mice. bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

A summary of all these bioenergetics-related changes that we observe in the EC of aged APOE4

mice is depicted in Fig. 5. Taken together, the observed differences on mitochondrial functionality among

the brain regions studied suggests that the EC possesses a different subset of compensatory mechanisms in

order to improve OxPhos efficiency under metabolic stress. While this apparent compensation of EC

mitochondria may be regarded as a positive attribute, it is plausible that it may have negative consequences

in the long run. For example, it has been shown that, over time, elevated mitochondrial respiratory activity

and increased coupling efficiency can result in an excess of free radical production and resulting oxidative

damage within cells (74).

More work will be necessary in order to further elucidate this phenomenon, including an

understanding of the specific contributions of different cell-types such as neurons, astrocytes and microglia.

It will also be important to investigate whether the other phenotypes that we have observed in the EC of

these aged APOE4 mice, such as endosomal-lysosomal dysregulation and neuronal hyperactivity, are

related to the bioenergetic observations described here. For example, Zhao et al. recently reported that

APOE4 expression causes insulin receptors to be inefficiently recycled via the endosomal-lysosomal

system, potentially causing the insulin resistance that the authors observed in the brains of aged APOE4

mice (25). As the endosomal-lysosomal dysregulation that we observed in the brains of aged APOE4 mice

was not restricted to the EC (41), this may play a role in the APOE4-associated bioenergetic deficits that

we observed throughout the brains of these mice. On the other hand, the neuronal hyperactivity that we

observed did appear to be primarily localized to the EC within the hippocampal formation (42). Intriguingly,

neuronal activity and glucose utilization have previously been shown to exist in a near stoichiometric

relationship (75), suggesting that the bioenergetic counterbalancing that we are observing in the EC of aged

APOE4 mice may be directly correlated to the observed neuronal hyperactivity in these mice (42).

Finally, it will be important to determine whether these bioenergetic differences in the EC are

involved in the increased risk of AD among APOE4 carriers. The EC is a unique brain region in many ways.

It is vital for learning and memory (especially spatial memory), acting as a bridge between the hippocampus

and the neocortex (76, 77). As such, it is known to have higher energy demands and to be more bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

metabolically active that many other brain regions (78). In AD, the EC is one of the first regions to develop

tauopathy, which eventually spreads into the hippocampus and other cortical regions as the disease

progresses (33). It is possible, therefore, that due to its unique bioenergetic needs, the EC may possess

additional mechanisms for increasing its bioenergetic capacity under conditions of metabolic stress such as

APOE4 expression. However, perhaps in the process of inducing these mechanisms, other negative

consequences may result, such as increased ROS production or increased tau hyperphosphorylation or

increased neuronal hyperactivity, such that it results in an earlier onset of AD.

METHODS

Mice

Human APOE targeted replacement mice were first developed by Sullivan et al. (79, 80) and were acquired

from Taconic Biosciences. All mice used in this study were treated in accordance with the National

Institutes of Health Guide for the Care and Use of Laboratory Animals and approved by the Columbia

University Medical Center and SUNY Medical Center Institutional Animal Care and Use Committee

(IACUC).

RNA-Sequencing

Male mice expressing human APOE3/3 (10 mice) or APOE3/4 (19 mice) were aged to 14-15 months, at

which point they were sacrificed by cervical dislocation, and brain tissues containing the EC and PVC were

dissected and snap frozen on dry-ice. Brain tissues were stored in RNase-free eppendorf tubes at -80°C

prior to extraction. Total RNA was extracted from frozen tissues by homogenizing each tissue sample using

a battery-operated pestle mixer (Argos Technologies, Vernon Hills, IL) in 1 ml of TRIzol reagent according

to the manufacturer's protocol (Life Technologies, Carlsbad, CA). RNA concentration was measured using

a nanodrop 1000 (Thermo Fisher Scientific, Waltham, MA), and RNA integrity (RIN) was assessed using

an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA). All RNA samples possessed a RIN bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

of 8 or higher. RNA was stored at -80˚C prior to use. Starting with 2 μg of total RNA per sample, Poly(A)+

mRNA was purified, fragmented and then converted into cDNA using the TruSeq RNA Sample Prep Kit

v2 (Illumina cat# RS-122-2001) as per the manufacturer’s protocol (Illumina, San Diego, CA). For RNA-

Sequencing of the cDNA, we hybridized 5 pM of each library to a flow cell, with a single lane for each

sample, and we used an Illumina cluster station for cluster generation. We then generated 149bp single end

sequences using an Illumina HiSeq 2000 sequencer. For analysis, we used the standard Illumina pipeline

with default options to analyze the images and extract base calls in order to generate fastq files. We then

aligned the fastq files to the mm9 mouse reference genome using Tophat (v2.0.6) and Bowtie (v2.0.2.0). In

order to annotate and quantify the reads to specific genes, we used the Python module HT-SEQ with a

modified NCBIM37.61 (containing only protein coding genes) gtf to provide reference boundaries. We

used the R/Bioconductor package DESeq2 (v1.10.1) for comparison of aligned reads across the samples.

We conducted a variance stabilizing transformation on the aligned and aggregated counts, and then the

Poisson distributions of normalized counts for each transcript were compared across APOE3/4 vs. APOE3/3

groups using a negative binomial test. We corrected for multiple testing using the Benjamini-Hochberg

procedure and selected all genes that possessed a corrected p-value of less than 0.05. Finally, a heat map

based on sample distance and a volcano plot based on fold change and adjusted p-values were generated

using the R/Bioconductor package ggplot2 (v2.0.0). For pathway analysis, enriched KEGG pathways were

identified using the ClueGo application (version 2.1.7) in Cytoscape (version 3.2.1). Briefly, all

differentially expressed EC genes from the RNA-Seq analysis were entered into the application and

searched for significantly enriched KEGG pathways possessing a Benjamini-Hochberg adjusted p-value of

less than 0.05.

Metabolomic Analysis

Male mice expressing human APOE3/3 (8 mice), APOE3/4 (9 mice), or APOE4/4 (7 mice) were aged to

14-15 months, at which point they were sacrificed by cervical dislocation to maintain the brain environment,

and individual brain regions were immediately removed and snap-frozen on dry ice. Tissues were stored at bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

-80°C for prior to extraction. Metabolite extraction was performed using a methyl tert-butyl ether

(MTBE)/methanol extraction protocol modified from previous reports (39, 40). Briefly, individual EC or

PVC tissues were homogenized in 400 l of ice-cold methanol using a bead mill homogenizer (TissueLyser

II, Qiagen) at 25 beats/sec, 2x for 45 sec each. Following homogenization, samples were incubated in 1200

l of MTBE for 1 hr at room temperature to separate organic-soluble lipids from aqueous-soluble lipids

and other small-molecules. Finally, 360 l of ultrapure water was added (for a final ratio of 3:1:0.9

MTBE:methanol:water) to resolve the two liquid phases, and each samples were centrifuged at 10,000 x g

for 10 min. The lower aqueous phase and the upper organic phase was collected from each sample and

stored in separate tubes, and the remaining protein pellets were resuspended in 25 mM ammonium

bicarbonate, pH 8, with 2.5% SDS. A BCA protein assay was performed on each protein fraction, and both

the aqueous phase and organic phase were normalized to their protein concentration equivalent with 50%

and 100% methanol respectively. All samples were then stored at -80°C prior to analysis. Metabolite

profiling was performed using an Agilent Model 1200 liquid chromatography (LC) system coupled to an

Agilent Model 6230 time-of-flight (TOF) mass analyzer as described previously (81). Metabolite separation

was accomplished using aqueous neutral phase (ANP) gradient chromatography on a Diamond Hydride

column (Microsolv), with mobile phase buffer A consisting of 50% isopropanol and 0.025% acetic acid,

and mobile phase buffer B consisting of 90% acetonitrile and 5 mM ammonium acetate. Each aqueous

phase sample was analyzed in both positive and negative ion detection mode for comprehensive coverage

of metabolites with a mass of 50-1000 Da. Prior to analysis, it was observed that one APOE3/4 PVC sample

possessed a chromatographic error and was removed from all future analysis.

Following MS analysis, the raw data was analyzed using Agilent Profinder (version B.06.00) and

Mass Profiler Professional (version 12.0) software. Briefly, Profinder performs batch recursive feature

extraction, which consists of a first-pass identification of features (ions) that possess a common elution

profile (e.g. identical mass-to-charge [m/z] ratios and retention times), followed by a recursive analysis to

re-mine each sample for the presence of features that were identified in at least one sample. Feature bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

extraction was limited to the top 2000 metabolites per ion detection mode, with each metabolite possessing

a minimum abundance of 600 ion counts. Each feature was also manually validated following extraction

(with investigators blinded to genotype assignments during validation), and re-integration was performed

when necessary. In addition to this untargeted feature extraction, we also used ProFinder to perform a

targeted feature extraction, matching features against a proprietary database of 626 biologically-relevant

metabolites whose mass and retention times were previously determined using pure chemical standards.

For this analysis, we allowed for a maximum mass deviation of 10 mDa and a retention time deviation of

1 min, and identified features were manually validated following extraction. For both untargeted and

targeted feature extraction, Profinder analysis was followed up by multivariate and differential expression

analysis using Mass Profiler Professional. Following differential expression analysis, metabolites identified

as differentially expressed were further validated for chromatographic and biological accuracy, and non-

validated metabolites were removed.

Seahorse Analysis

In order to investigate the effect of differential apoE isoform expression on electron transport chain activity,

mitochondria were isolated from tissues collected from 12-17 month old apoE3/3 or E4/4 mice (10 mice

per group, separated into three experiments), and the oxygen consumption rates (OCR) were measured in

a complex I or complex II-mediated fashion, as previously described (49). Briefly, each mouse was

sacrificed by cervical dislocation, and the different brain areas from each brain hemisphere were

immediately removed and placed into ice-cold PBS containing protease inhibitors. For each experiment,

tissues were pooled from 3-4 mice per apoE group. Following tissue collection, samples were homogenized

in ~10 volumes of isolation buffer (70 mM sucrose, 210 mM mannitol, 5 mM HEPES, 1 mM EGTA and

0.5% fatty acid-free BSA, pH 7.2) and rinsed 3 times. Tissues were then homogenized using a Teflon glass

homogenizer with 2-3 strokes. Homogenized samples were then centrifuged at 800 x g for 10 min at 4°C.

Following centrifugation, fat/lipid was aspirated off, and the remaining supernatant was decanted through

2 layers of cheesecloth into a separate tube and centrifuged at 8000 x g for 10 min at 4°C. After removal of bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

the supernatant, the pellet was resuspended in mitochondrial isolation buffer, and the centrifugation was

repeated. The resulting pellet was resuspended in mitochondrial isolation buffer, and total protein

concentrations were determined using Bradford Assay reagent (Bio-Rad).

For complex I experiments, 8 g of protein was added to each well and for complex II analysis 6 g per well. To prepare samples for OCR measurements, samples were prepared in mitochondrial assay buffer (70 mM

sucrose, 220 mM mannitol, 10 mM KH2PO4, 5 mM MgCl2, 2 mM HEPES, 1 mM EGTA and 0.2% fatty

acid-free BSA, pH 7.2) plus substrate (pyruvate and malate for complex I-mediated ETC activity assay, or

succinate and rotenone for complex II-mediated ETC activity assay).

Respirometry was performed using the XF24e Extracellular Flux Analyzer (Seahorse Bioscience). Oxygen consumption was measured in basal conditions (Seahorse media with 25 mM glucose and 2 mM pyruvate) and after the sequential addition of 1 μM oligomycin (Complex V inhibitor), 0.75 μM FCCP (uncoupler) and 1 μM rotenone/1 μM antimycin A (Complex I and Complex III inhibitors respectively). All results are averages of five or more biological replicates. Every biological replicate consisted of three technical replicates. OCR data was generated by the Seahorse XF24 1.5.0.69 software package and displayed in point-to-point

mode. Final calculations of total OCR were performed by normalizing apoE4/4 values to apoE3/3 for each

of the three experiments and then combining the data in Prism.

PIUMet Analysis

We used the previously published PIUMet algorithm (51) to analyze the untargeted metabolomics data

from APOE4/4and APOE3/3 in the EC of aged APOE mice. PIUMet discovers dysregulated molecular

pathways and components associated with changes in untargeted metabolomic data by analyzing a network

built from curated protein-protein and protein-metabolite interactions (PPMI). PIUMet represents each

metabolite peak as a node in this network and connects it to metabolites with masses that correspond to the

m/z values of the peak. Using the prize-collecting Steiner tree algorithm, it searches the PPMI interactome

to find an optimum subnetwork that connects the input metabolite peaks via their putative identities and

other metabolites and proteins that were not detected in the experiments. The Gene Ontology (GO) analysis bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

of the resulting network further reveals associated molecular pathways with the APOE4 phenotype. We

selected peaks 304 untargeted metabolite peaks or features that were significantly altered between

APOE4/4and APOE3/3 models (corrected p-value <0.05). We then assigned a prize to each input data point

to show the significance of their alterations, as –log (P values) of the significance of changes between two

phenotypic conditions. PIUMet accepts several parameters that regulate the size of the resulting networks,

including w and beta. While w tunes the number of trees in the resulting network, beta tunes the number of

input nodes that are included in the output. Another parameter, mu, controls the bias toward high degree

nodes. Higher values of mu results in a lower number of high degree nodes in the resulting networks, and

thus less bias to highly studied molecules or ubiquitous interactions such as those with ions. We determined

the optimum parameters (w=6.0, beta=0.5, and mu=0.0005) based on their effects on the size of the resulting

networks. We further generated 100 networks by adding random noise to the underlying database, and

calculated robustness scores for each node. Nodes with robustness scores less than 60% were removed from

the results. Additionally, for each node we calculated a specificity score, based on the number of their

presence in the 100 random networks obtained from a set of mock metabolite peaks randomly selected from

the input feature list that mimic real input. A score of 100 indicates that the node did not show in any of the

randomly generated networks, while a zero score shows the node appears in all the randomly generated

networks. 79% of the resulting nodes have specificity score of over 60%.

1H NMR Spectroscopy

All animal procedures for in vivo proton MR spectroscopy were performed following the National Institutes

of Health guidelines with approval from the Institutional Animal Care and Use Committee at the Nathan S.

Kline Institute for Psychiatric Research. All animals were anesthetized using an isoflurane vaporizer set at

the following percentages: 3% for induction, 2% during pilot scanning and 1.5% during data acquisition.

An animal monitoring unit (model 1025, SA Instruments, Stony Brook, NY, USA) was used to record

respiration and rectal temperature. Respiration was measured with a pressure transducer placed under the

abdomen just below the ribcage. Body temperature was maintained using forced warm air, controlled by a bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

feedback circuit between the heater and thermistor. After induction, the animals were placed on a holder

and restrained using a bite bar and ear bars placed half way into the auditory canal. Oxygen was used as the

carrier gas delivered to a cone positioned before the bite bar, where gases mixed with air and passed over

the rodent’s nose. All animals were maintained at 37.0 ± 0.2 °C. All data were obtained on a 7.0 T Agilent

(Santa Clara, CA, USA) 40 cm bore system. The gradient coil insert had an internal diameter of 12 cm with

a maximum gradient strength of 600 mT/m and minimum rise time of 200 μs, with customized second and

third order shim coils. A Rapid (Rimpar, Germany) volume transmit coil (72mm ID) and a two-channel

receive-only surface coil was used for RF transmission and reception, respectively.

The shim settings for the selected volume of interest (VOI) were automatically adjusted using

FASTMAP1 (Fast, Automatic Shimming Technique by Mapping Along Projections), a high order shim

method, which samples the magnetic field along a group of radial columns which focus on the center of a

localized voxel. It is a method for optimizing the field homogeneity in a cubical local region.

The water signal was suppressed using variable power RF pulses with optimized relaxation delays

(VAPOR)2. The spectral acquisition consisted of a short echo time Point Resolved Spectroscopy (PRESS)3

sequence with the following parameters: repetition time = 4s, echo time = 7.5 ms, number of averages =

512 acquired in blocks of 128, number of points = 2048 and bandwidth of acquisition = 5 kHz, total

acquisition time = 34 minutes. Outer volume suppression was also used to minimize signal contamination

by extra cranial muscle and lipids. The VOI size was 3.3 µl (2.0x1.3x1.3 mm3) placed in the entorhinal

cortex (EC). The target VOI is depicted in Fig. 4 overlaid on a schematic of the coronal brain slice at

approximately – 2.92 mm relative to Bregma. An anatomical T2 weighted pilot scan was used to position

the VOI (coronal). These scans were acquired with a fast spin echo sequence with the following parameters:

field of view = 20 mm with 256x256 matrix size, slice thickness = 0.5 mm, number of slices = 11, repetition

time = 4s, echo train length = 8, echo spacing = 15 ms, effective echo time = 60 ms, number of averages =

8, total acquisition time = 8 minutes 40 s. All data were processed using the LCModel software developed

by Provencher 4. This software calculates the best fit to the acquired data of a linear combination of model

spectra acquired from in vitro solutions of all the brain metabolites of interest. This basis set of the model bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

spectra has the same echo time, sequence acquisition and field strength as the acquired data of the study.

The LCModel software outputs the estimated concentration along with estimated standard deviations

(Cramer-Rao lower bounds) expressed in percent of the estimated concentration, which can be used as a

quantitative measure of reliability. To improve statistical significance, Miller et al. recently demonstrated

the use of weighted averaging in NMR spectroscopic studies (82). In this current study, we also used

weighted averaging to calculate the standard unequal variance t-test (Welch’s t-test) as outlined by Miller

(82) and in the reference manual of the LCModel software (83). All data used in the final analysis had

Cramer-Rao lower bounds of 20% or less. The metabolites measured were: alanine( Ala), aspartate (Asp),

creatine (Cr), phosphocreatine (PCr), ɣ-Aminobutyric Acid (GABA), glucose (Glc), glutamine (Gln),

glutamate (Glu), glycerophosphocholine (GPC), phosphocholine (PCh), glutathione (GSH), myo-inositol

(Ins), lactate (Lac), N-Acetylaspartate (NAA), N-Acetylaspartateglutamate (NAAG) and taurine (Tau). It

is often quite difficult to resolve Glu from Gln, NAA from NAAG and PCh from GPC, particularly if the

spectral quality is poor due to an inadequate shim. So, in addition to the principal metabolites the total of

Glu and Gln, NAA and NAAG and PCh and GPC are also reported. These total concentrations are thought

to be a more reliable metric. An unsuppressed water signal was also used for absolute concentration

calculation and eddy current correction. This internal reference method assumes known values of water

concentrations of gray and white matter5-7.

There are several limitations to 1H NMR spectroscopy which can make reliable measures

particularly challenging. One such limitation is that is that 1H NMR metabolite measures are inherently low

in signal to noise. The metabolite signals measured are approximately 10,000-fold less than the proton

signal used in imaging. The EC is also a small structure and thus requires a small VOI which reduces the

measured signal strength. However, at 7T with shim values in the range 8 – 15 Hz and a short echo time of

8 ms, the spectral peaks were reasonably separated, allowing for direct and reliable measures of metabolite

concentrations. For metabolites which have significant spectral overlap, total values (e.g. Glu and Gln) are

also reported. Another error source is chemical shift displacement. The bandwidth of the 180º RF pulse is

5 kHz, which would give a displacement error of 0.3 mm in the vertical and horizontal directions over the bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

range of 0.2 to 4.2 ppm. However, this displacement was relatively small and still allowed for sufficient

specificity. Despite the challenges of in vivo mouse brain 1H NMR, we were able to achieve high quality

spectra in this difficult region with high quantification accuracy.

Data Availability

The transcriptomics datasets are available in the NCBI Gene Expression Omnibus (GEO)

repository: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE102334. The metabolomics

datasets are available in the MetaboLights repository: http://www.ebi.ac.uk/metabolights/MTBLS530

ACKNOWLEDGEMENTS

We acknowledge Drs. Wai-Haung Yu and Natura Myeku for advice on this manuscript, as well as Laidon

Piroli for assistance with genotyping. We also acknowledge Dr. Paul Mathews for providing the APOE

mice that were used in the 1H NMR experiment. This work was supported by grants from NIA and NINDS

to K.E.D. (AG048408 and NS071836), grants from NIA to E.A.G. (AG045335 and AG056387), grants

from NICHD and NHLBI to S.S.G. (HD67244 and HL87062), grants from NIA and NINDS to E.F.

(AG057331 and NS089076), a grant from NCCR (1S10RR023534-01) for the 7T rodent scanner at Nathan

Kline Institute, grants from the U.S. Dept. of Defense to D.L., (W911NF-12-1-9159 and W911F-15-1-

0169), funding from the Cure Alzheimer's Fund to KD and TN, and fellowships from NIA (AG047797)

and the Brightfocus Foundation to T.N. In addition, this work was supported in part by the Intramural

Research Program of the NIH, National Institute on Aging.

AUTHOR CONTRIBUTIONS bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

This study was designed and managed by T.N., E.A.G., and K.E.D. Animal care and breeding was

performed by H.F. Transcriptomic analysis was performed by A.A.D., in the laboratory of M.R.C.

Metabolomic analysis was performed by Q.C. and T.N., in the laboratory of S.S.G. Seahorse analysis was

performed by D.L. and M.P., in the laboratory of E.A.G. PIUMet analysis was performed by L.P., in the

laboratory of E.F. 1H NMR spectroscopy was performed by K.S. and D.G. Additional molecular biology

experiments and bioinformatics were performed by T.N., A.A. and H.A. Manuscript preparation was

performed by T.N. and E.A.G., with input from K.E.D.

COMPETING FINANCIAL INTERESTS

K.E.D is on the board of directors of Ceracuity LLC. The authors declare no competing financial interests.

bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

REFERENCES

1. R. W. Mahley, S. C. Rall, Jr., Apolipoprotein E: far more than a lipid transport protein. Annual review of genomics and human genetics 1, 507-537 (2000). 2. X. Han, The role of apolipoprotein E in lipid metabolism in the central nervous system. Cell Mol Life Sci 61, 1896-1906 (2004). 3. D. M. Holtzman, J. Herz, G. Bu, Apolipoprotein E and apolipoprotein E receptors: normal biology and roles in Alzheimer disease. Cold Spring Harbor perspectives in medicine 2, a006312 (2012). 4. C. C. Liu, T. Kanekiyo, H. Xu, G. Bu, Apolipoprotein E and Alzheimer disease: risk, mechanisms and therapy. Nat Rev Neurol 9, 106-118 (2013). 5. J. Davignon, R. E. Gregg, C. F. Sing, Apolipoprotein E polymorphism and atherosclerosis. Arteriosclerosis 8, 1-21 (1988). 6. A. M. Cumming, F. W. Robertson, Polymorphism at the apoprotein-E locus in relation to risk of coronary disease. Clin Genet 25, 310-313 (1984). 7. A. M. Saunders et al., Association of apolipoprotein E allele epsilon 4 with late-onset familial and sporadic Alzheimer's disease. Neurology 43, 1467-1472 (1993). 8. W. J. Strittmatter et al., Apolipoprotein E: high-avidity binding to beta-amyloid and increased frequency of type 4 allele in late-onset familial Alzheimer disease. Proc Natl Acad Sci U S A 90, 1977-1981 (1993). 9. L. A. Farrer et al., Effects of age, sex, and ethnicity on the association between apolipoprotein E genotype and Alzheimer disease. A meta-analysis. APOE and Alzheimer Disease Meta Analysis Consortium. JAMA 278, 1349-1356 (1997). 10. S. B. Sando et al., APOE epsilon 4 lowers age at onset and is a high risk factor for Alzheimer's disease; a case control study from central Norway. BMC neurology 8, 9 (2008). 11. D. C. Rubinsztein, D. F. Easton, Apolipoprotein E genetic variation and Alzheimer's disease. a meta-analysis. Dementia and geriatric cognitive disorders 10, 199-209 (1999). 12. K. R. Bales et al., Lack of apolipoprotein E dramatically reduces amyloid beta- peptide deposition. Nat Genet 17, 263-264 (1997). 13. E. M. Castano et al., Fibrillogenesis in Alzheimer's disease of amyloid beta peptides and apolipoprotein E. Biochem J 306 ( Pt 2), 599-604 (1995). 14. G. W. Rebeck, J. S. Reiter, D. K. Strickland, B. T. Hyman, Apolipoprotein E in sporadic Alzheimer's disease: allelic variation and receptor interactions. Neuron 11, 575-580 (1993). 15. D. E. Schmechel et al., Increased amyloid beta-peptide deposition in cerebral cortex as a consequence of apolipoprotein E genotype in late-onset Alzheimer disease. Proc Natl Acad Sci U S A 90, 9649-9653 (1993). 16. J. Ma, A. Yee, H. B. Brewer, Jr., S. Das, H. Potter, Amyloid-associated proteins alpha 1- antichymotrypsin and apolipoprotein E promote assembly of Alzheimer beta- protein into filaments. Nature 372, 92-94 (1994). 17. J. M. Castellano et al., Human apoE isoforms differentially regulate brain amyloid- beta peptide clearance. Science translational medicine 3, 89ra57 (2011). bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

18. D. M. Holtzman et al., Apolipoprotein E isoform-dependent amyloid deposition and neuritic degeneration in a mouse model of Alzheimer's disease. Proc Natl Acad Sci U S A 97, 2892-2897 (2000). 19. Y. Huang, Abeta-independent roles of apolipoprotein E4 in the pathogenesis of Alzheimer's disease. Trends in molecular medicine 16, 287-294 (2010). 20. A. B. Wolf et al., Apolipoprotein E as a beta-amyloid-independent factor in alzheimer's disease. Alzheimers Res Ther 5, 38 (2013). 21. E. M. Reiman et al., Preclinical evidence of Alzheimer's disease in persons homozygous for the epsilon 4 allele for apolipoprotein E. The New England journal of medicine 334, 752-758 (1996). 22. E. M. Reiman et al., Functional brain abnormalities in young adults at genetic risk for late-onset Alzheimer's dementia. Proc Natl Acad Sci U S A 101, 284-289 (2004). 23. E. M. Reiman et al., Correlations between apolipoprotein E epsilon4 gene dose and brain-imaging measurements of regional hypometabolism. Proc Natl Acad Sci U S A 102, 8299-8302 (2005). 24. Q. R. Ong, E. S. Chan, M. L. Lim, G. M. Cole, B. S. Wong, Reduced phosphorylation of brain insulin receptor substrate and Akt proteins in apolipoprotein-E4 targeted replacement mice. Scientific reports 4, 3754 (2014). 25. N. Zhao et al., Apolipoprotein E4 Impairs Neuronal Insulin Signaling by Trapping Insulin Receptor in the Endosomes. Neuron 96, 115-129 e115 (2017). 26. L. A. Johnson et al., Apolipoprotein E4 mediates insulin resistance-associated cerebrovascular dysfunction and the post-prandial response. Journal of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism, 271678X17746186 (2017). 27. L. A. Johnson, E. R. Torres, S. Impey, J. F. Stevens, J. Raber, Apolipoprotein E4 and Insulin Resistance Interact to Impair Cognition and Alter the Epigenome and Metabolome. Scientific reports 7, 43701 (2017). 28. J. T. Keeney, S. Ibrahimi, L. Zhao, Human ApoE Isoforms Differentially Modulate Glucose and Amyloid Metabolic Pathways in Female Brain: Evidence of the Mechanism of Neuroprotection by ApoE2 and Implications for Alzheimer's Disease Prevention and Early Intervention. Journal of Alzheimer's disease : JAD 48, 411-424 (2015). 29. L. Wu, X. Zhang, L. Zhao, Human ApoE Isoforms Differentially Modulate Brain Glucose and Ketone Body Metabolism: Implications for Alzheimer's Disease Risk Reduction and Early Intervention. The Journal of neuroscience : the official journal of the Society for Neuroscience 38, 6665-6681 (2018). 30. S. Chang et al., Lipid- and receptor-binding regions of apolipoprotein E4 fragments act in concert to cause mitochondrial dysfunction and neurotoxicity. Proc Natl Acad Sci U S A 102, 18694-18699 (2005). 31. Y. Huang et al., Apolipoprotein E fragments present in Alzheimer's disease brains induce neurofibrillary tangle-like intracellular inclusions in neurons. Proc Natl Acad Sci U S A 98, 8838-8843 (2001). 32. T. Nakamura, A. Watanabe, T. Fujino, T. Hosono, M. Michikawa, Apolipoprotein E4 (1-272) fragment is associated with mitochondrial proteins and affects mitochondrial function in neuronal cells. Mol Neurodegener 4, 35 (2009). bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

33. H. Braak, E. Braak, Neuropathological stageing of Alzheimer-related changes. Acta neuropathologica 82, 239-259 (1991). 34. M. Bobinski et al., MRI of entorhinal cortex in mild Alzheimer's disease. Lancet 353, 38-40 (1999). 35. P. Shaw et al., Cortical morphology in children and adolescents with different apolipoprotein E gene polymorphisms: an observational study. Lancet neurology 6, 494-500 (2007). 36. G. A. Rodriguez, M. P. Burns, E. J. Weeber, G. W. Rebeck, Young APOE4 targeted replacement mice exhibit poor spatial learning and memory, with reduced dendritic spine density in the medial entorhinal cortex. Learning & memory 20, 256-266 (2013). 37. X. Wang, M. L. Michaelis, E. K. Michaelis, Functional genomics of brain aging and Alzheimer's disease: focus on selective neuronal vulnerability. Current genomics 11, 618-633 (2010). 38. S. Minoshima et al., Metabolic reduction in the posterior cingulate cortex in very early Alzheimer's disease. Ann Neurol 42, 85-94 (1997). 39. P. Giavalisco et al., Elemental formula annotation of polar and lipophilic metabolites using (13) C, (15) N and (34) S isotope labelling, in combination with high- resolution mass spectrometry. Plant J 68, 364-376 (2011). 40. V. Matyash, G. Liebisch, T. V. Kurzchalia, A. Shevchenko, D. Schwudke, Lipid extraction by methyl-tert-butyl ether for high-throughput lipidomics. J Lipid Res 49, 1137-1146 (2008). 41. T. Nuriel et al., The Endosomal-Lysosomal Pathway Is Dysregulated by APOE4 Expression in Vivo. Front Neurosci 11, 702 (2017). 42. T. Nuriel et al., Neuronal hyperactivity due to loss of inhibitory tone in APOE4 mice lacking Alzheimer's disease-like pathology. Nature communications 8, 1464 (2017). 43. D. Speijer, Oxygen radicals shaping evolution: why fatty acid catabolism leads to peroxisomes while neurons do without it: FADH(2)/NADH flux ratios determining mitochondrial radical formation were crucial for the eukaryotic invention of peroxisomes and catabolic tissue differentiation. BioEssays : news and reviews in molecular, cellular and developmental biology 33, 88-94 (2011). 44. S. P. Wang et al., Metabolism as a tool for understanding human brain evolution: lipid energy metabolism as an example. Journal of human evolution 77, 41-49 (2014). 45. P. Schonfeld, G. Reiser, Why does brain metabolism not favor burning of fatty acids to provide energy? Reflections on disadvantages of the use of free fatty acids as fuel for brain. J Cereb Blood Flow Metab 33, 1493-1499 (2013). 46. K. A. Nalecz, D. Miecz, V. Berezowski, R. Cecchelli, Carnitine: transport and physiological functions in the brain. Molecular aspects of medicine 25, 551-567 (2004). 47. D. Ebert, R. G. Haller, M. E. Walton, Energy contribution of octanoate to intact rat brain metabolism measured by 13C nuclear magnetic resonance spectroscopy. J Neurosci 23, 5928-5935 (2003). 48. A. Panov, Z. Orynbayeva, V. Vavilin, V. Lyakhovich, Fatty acids in energy metabolism of the central nervous system. BioMed research international 2014, 472459 (2014). bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

49. G. W. Rogers et al., High throughput microplate respiratory measurements using minimal quantities of isolated mitochondria. PLoS One 6, e21746 (2011). 50. M. D. Brand, D. G. Nicholls, Assessing mitochondrial dysfunction in cells. Biochem J 435, 297-312 (2011). 51. L. Pirhaji et al., Revealing disease-associated pathways by network integration of untargeted metabolomics. Nature methods 13, 770-776 (2016). 52. A. M. Robinson, D. H. Williamson, Physiological roles of ketone bodies as substrates and signals in mammalian tissues. Physiological reviews 60, 143-187 (1980). 53. Y. Zhang et al., Ketosis proportionately spares glucose utilization in brain. Journal of cerebral blood flow and metabolism : official journal of the International Society of Cerebral Blood Flow and Metabolism 33, 1307-1311 (2013). 54. A. L. Lin, W. Zhang, X. Gao, L. Watts, Caloric restriction increases ketone bodies metabolism and preserves blood flow in aging brain. Neurobiology of aging 36, 2296-2303 (2015). 55. J. Thevenet et al., Medium-chain fatty acids inhibit mitochondrial metabolism in astrocytes promoting astrocyte-neuron lactate and ketone body shuttle systems. FASEB journal : official publication of the Federation of American Societies for Experimental Biology 30, 1913-1926 (2016). 56. N. Auestad, R. A. Korsak, J. W. Morrow, J. Edmond, Fatty acid oxidation and ketogenesis by astrocytes in primary culture. Journal of neurochemistry 56, 1376- 1386 (1991). 57. M. Guzman, C. Blazquez, Ketone body synthesis in the brain: possible neuroprotective effects. Prostaglandins, leukotrienes, and essential fatty acids 70, 287-292 (2004). 58. M. G. Bixel, B. Hamprecht, Generation of ketone bodies from leucine by cultured astroglial cells. Journal of neurochemistry 65, 2450-2461 (1995). 59. R. Murin, B. Hamprecht, Metabolic and regulatory roles of leucine in neural cells. Neurochemical research 33, 279-284 (2008). 60. A. H. Emwas, The strengths and weaknesses of NMR spectroscopy and mass spectrometry with particular focus on metabolomics research. Methods in molecular biology 1277, 161-193 (2015). 61. S. P. Bessman, The creatine phosphate energy shuttle--the molecular asymmetry of a "pool". Analytical biochemistry 161, 519-523 (1987). 62. W. Chen, X. H. Zhu, G. Adriany, K. Ugurbil, Increase of creatine kinase activity in the visual cortex of human brain during visual stimulation: a 31P magnetization transfer study. Magnetic resonance in medicine 38, 551-557 (1997). 63. A. Schousboe, L. K. Bak, H. S. Waagepetersen, Astrocytic Control of Biosynthesis and Turnover of the Neurotransmitters Glutamate and GABA. Frontiers in endocrinology 4, 102 (2013). 64. M. Kreft, L. K. Bak, H. S. Waagepetersen, A. Schousboe, Aspects of astrocyte energy metabolism, amino acid neurotransmitter homoeostasis and metabolic compartmentation. ASN neuro 4, (2012). 65. A. Falkowska et al., Energy Metabolism of the Brain, Including the Cooperation between Astrocytes and Neurons, Especially in the Context of Glycogen Metabolism. International journal of molecular sciences 16, 25959-25981 (2015). bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

66. A. B. Patel et al., The contribution of GABA to glutamate/glutamine cycling and energy metabolism in the rat cortex in vivo. Proc Natl Acad Sci U S A 102, 5588-5593 (2005). 67. R. Rossignol, M. Malgat, J. P. Mazat, T. Letellier, Threshold effect and tissue specificity. Implication for mitochondrial cytopathies. J Biol Chem 274, 33426-33432 (1999). 68. A. M. Celotto, W. K. Chiu, W. Van Voorhies, M. J. Palladino, Modes of metabolic compensation during mitochondrial disease using the Drosophila model of ATP6 dysfunction. PLoS One 6, e25823 (2011). 69. D. C. Chan, Mitochondria: dynamic organelles in disease, aging, and development. Cell 125, 1241-1252 (2006). 70. B. E. Sansbury, S. P. Jones, D. W. Riggs, V. M. Darley-Usmar, B. G. Hill, Bioenergetic function in cardiovascular cells: the importance of the reserve capacity and its biological regulation. Chemico-biological interactions 191, 288-295 (2011). 71. B. H. Goodpaster, L. M. Sparks, Metabolic Flexibility in Health and Disease. Cell metabolism 25, 1027-1036 (2017). 72. F. Palmieri, The mitochondrial transporter family SLC25: identification, properties and physiopathology. Molecular aspects of medicine 34, 465-484 (2013). 73. V. A. Saks et al., Control of cellular respiration in vivo by mitochondrial outer membrane and by creatine kinase. A new speculative hypothesis: possible involvement of mitochondrial-cytoskeleton interactions. Journal of molecular and cellular cardiology 27, 625-645 (1995). 74. I. Bratic, A. Trifunovic, Mitochondrial energy metabolism and ageing. Biochimica et biophysica acta 1797, 961-967 (2010). 75. G. Yellen, Fueling thought: Management of glycolysis and oxidative phosphorylation in neuronal metabolism. The Journal of cell biology 217, 2235-2246 (2018). 76. H. Eichenbaum, A cortical-hippocampal system for declarative memory. Nat Rev Neurosci 1, 41-50 (2000). 77. L. M. Frank, E. N. Brown, M. Wilson, Trajectory encoding in the hippocampus and entorhinal cortex. Neuron 27, 169-178 (2000). 78. A. V. Egorov, B. N. Hamam, E. Fransen, M. E. Hasselmo, A. A. Alonso, Graded persistent activity in entorhinal cortex neurons. Nature 420, 173-178 (2002). 79. P. M. Sullivan, B. E. Mace, N. Maeda, D. E. Schmechel, Marked regional differences of brain human apolipoprotein E expression in targeted replacement mice. Neuroscience 124, 725-733 (2004). 80. P. M. Sullivan et al., Targeted replacement of the mouse apolipoprotein E gene with the common human APOE3 allele enhances diet-induced hypercholesterolemia and atherosclerosis. J Biol Chem 272, 17972-17980 (1997). 81. Q. Chen et al., Untargeted plasma metabolite profiling reveals the broad systemic consequences of xanthine oxidoreductase inactivation in mice. PLoS One 7, e37149 (2012). 82. J. J. Miller, L. Cochlin, K. Clarke, D. J. Tyler, Weighted averaging in spectroscopic studies improves statistical power. Magnetic resonance in medicine 78, 2082-2094 (2017). 83. S. W. Provencher, Estimation of metabolite concentrations from localized in vivo proton NMR spectra. Magnetic resonance in medicine 30, 672-679 (1993). bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

FIGURE LEGENDS Fig. 1. Transcriptomics analysis reveals an upregulation of electron transport chain genes in the EC of aged APOE4 mice. RNA-sequencing was performed in order to analyze the effects of differential APOE isoform expression on RNA levels in the EC and PVC of aged APOE mice (19 APOE3/4 males vs. 10 APOE3/3 males). (A) Shown here are the significantly enriched KEGG pathways observed in the EC of APOE3/4 vs. APOE3/3 mice, using Cytoscape’s ClueGo application, with the ten differentially expressed electron transport chain (ETC) genes circled. (B) Graphs of the differentially regulated ETC genes in the EC of the APOE3/4 vs. APOE3/3 mice. (** denotes p < 0.01; *** denotes p < 0.001; **** denotes p < 0.0001)

Fig. 2. Metabolomics analysis reveals differential expression of energy-related metabolites and fatty acids in the EC of aged APOE4 mice. An untargeted metabolomics analysis was performed in order to analyze the effects of differential APOE isoform expression on small-molecule metabolite levels in the EC and PVC of aged APOE mice (8 APOE3/3, 9 APOE3/4 and 7 APOE4/4 males). (A-C) Shown here are graphs of the bioenergetics-related metabolite groups that were shown to be differentially expressed in the EC of the APOE4/4 vs. APOE3/3 mice: (A) General energy-related metabolites and (B) Fatty acids. (C) We have also calculated the ratio of ATP:ADP in each genotype group from the EC and PVC of the aged APOE mice. (* denotes p < 0.05; ** denotes p < 0.01; **** denotes p < 0.0001)

Fig. 3. Seahorse analysis reveals decreased mitochondrial respiration in the cortex and hippocampus, but not in the EC, of aged APOE4 mice. Seahorse analysis was performed in order to analyze the effects of differential APOE isoform expression on mitochondrial respiration in mitochondria that were purified from the cortex (Ctx), hippocampus (Hip) and EC of aged APOE mice (2 APOE4/4 males, tissues pooled vs. 2 APOE3/3 males, tissues pooled). (A) The complex I- and complex II-driven ETC activity from each region shows decreased mitochondrial respiration in the mitochondria purified from the cortex and hippocampus, but not the EC of the aged APOE4/4 mice. (B-C) Bar graphs showing the average oxygen consumption rate (OCR) from (B) State 3 and for (C) the Respiration Control Ratio (RCR; state 3u/state 4o) in mitochondria purified from each region of the APOE4/4 mice, as a percentage of the APOE3/3 OCR from the equivalent tissues. The dotted blue line represents the normalized levels in the APOE3/3 tissues. (* denotes p < 0.05; ** denotes p < 0.01; *** denotes p < 0.001; **** denotes p < 0.0001)

Fig. 4. Proton nuclear magnetic resonance spectroscopy reveals differential expression of phosphocreatine, glutamate and GABA in the EC of aged APOE4 mice. 1H NMR spectroscopy was performed in order to analyze the effects of differential APOE isoform expression on the levels of high- abundance metabolites in the EC of anesthetized APOE mice (7 APOE4/4 males vs. 7 APOE3/3 males). (A) A representative coronal anatomical image used in the analysis, with the volume of interest, highlighted in green, shown over the EC. (B) A typical spectrum from the EC showing the spectral position of some major metabolites: N-Acetylaspartic acid (NAA), glutamate (Glu), total creatine (tCr) and total choline (tCho). The raw spectrum is shown in black (with no smoothing), and the software fit of the model spectra shown in red. The top of the figure shows the difference between the raw spectrum and the fit, with a “good fit” showing random oscillations with little noise. The solid black line at bottom of the figure is the baseline. (C) The metabolite concentrations in the EC of aged APOE3/3 and APOE4/4 mice. The metabolites reported are: alanine (Ala), aspartate (Asp), creatine (Cr), phosphocreatine (PCr), gamma- aminobutyric acid (GABA), glucose (Glc), glutamine (Gln), glutamate (Glu), glycerophosphocholine (GPC), phosphocholine (PCh), glutathione (GSH), myo-inositol (Ins), lactate (Lac), N-Acetylaspartate (NAA), N-Acetylaspartateglutamate (NAAG), taurine (Tau), total choline (GPC and PCh), total NAA (NAA and NAAG), total creatine (Cr and PCr) and total glutamate and glutamine (Glu and Gln). (* denotes p < 0.05; ** denotes p < 0.01; *** denotes p < 0.001; **** denotes p < 0.0001) bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Fig. 5. Identification of differentially expressed bioenergetic genes and metabolites in the EC of aged APOE4 mice. A schematic of the bioenergetics genes and metabolites that have been predicted to be altered by differential APOE isoform expression in this study, as depicted in a simplified manner with respect to their relationships to each other and their location inside or outside of mitochondria. Genes or metabolites that were putatively identified as increased in aged APOE4 mice, as compared to aged APOE3 mice, are marked with a green arrow, and those that are decreased are marked with a red arrow.

Fig. 6. Targeted lipidomics reveals differential expression of diacyglycerols and bis(monacylglycero)phosphates in the EC of aged APOE4 mice. A targeted lipidomic analysis was performed in order to analyze the effects of differential APOE isoform expression on lipid levels in the EC and PVC of aged APOE mice (8 APOE3/3, 9 APOE3/4 and 7 APOE4/4 males). (A-C) Shown here are graphs of the three primary groups of lipids that were shown to be differentially expressed in the EC of the APOE4/4 vs. APOE3/3 mice: (A) Diacylglycerols, (B) Bis(monoacylglycero)phosphates, and (C) Hexosylceramides, which include both glucosylceramides and galactosylceramides. (* denotes p < 0.05; ** denotes p < 0.01; *** denotes p < 0.001; **** denotes p < 0.0001)

bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Fig. 1

bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Fig. 2

bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Fig. 3

bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Fig. 4

bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Fig. 5

bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Table 1. Differentially expressed targeted metabolites from the EC of APOE3 vs. E4 mice

Regulation Fold Detection Metabolite p-value FDR CAS Number in E4/4 Change Mode Fatty Acids gamma-Linolenic Acid down 1.31 1.19E-03 0.093 positive 506-26-3 Myristic Acid down 3.92 0.004 0.118 positive 544-63-8 Docosahexaenoic Acid up 1.36 0.011 0.154 positive 6217-54-5 Stearic Acid down 1.69 0.011 0.154 positive 57-11-4 12-Hydroxydodecanoic Acid down 1.36 0.021 0.213 positive 505-95-3 Arachidic Acid down 1.40 0.037 0.230 negative 506-30-9 Palmitic Acid down 1.32 0.049 0.243 negative 57-10-3 Oligosaccharides trisaccharide up 1.99 0.001 0.049 negative 512-69-6 tetrasaccharide up 2.24 0.015 0.169 negative 10094-58-3 disaccharide up 2.20 0.015 0.169 negative 63-42-3 Vitamin and Vitamin

Derivatives Phylloquinone up 1.47 0.003 0.102 positive 84-80-0 Tocopherol up 2.74 0.005 0.123 negative 59-02-9 Dehydroascorbic acid up 1.76 0.008 0.134 positive 490-83-5 Energy-Related

Metabolites Inosine 5'-monophosphate 0.072 up 2.26 0.003 negative 131-99-7 (IMP) 0 D-Fructose 6-phosphate up 1.10 0.011 0.169 negative 643-13-0 Succinoadenosine up 1.47 0.011 0.169 negative 4542-23-8 Carnitine up 1.27 0.021 0.213 positive 541-15-1 77-92-9/ Citric Acid/Isocitric Acid up 1.24 0.028 0.230 negative 1637-73-6 Malic Acid up 1.16 0.037 0.230 negative 97-67-6 ATP up 1.24 0.049 0.243 negative 56-65-5 Cholesterol Metabolites Lanosterol up 2.27 0.015 0.169 negative 79-63-0 Cholesteryl Acetate up 2.29 0.015 0.169 negative 604-35-3 Amino Acids Leucine up 1.26 0.015 0.169 negative 61-90-5 Proline up 1.12 0.037 0.230 negative 147-85-3 Glycine up 1.10 0.037 0.230 negative 56-40-6 Tryptophan Metabolites Quinaldic Acid up 1.68 0.001 0.049 negative 93-10-7 Kynurenine up 2.49 0.015 0.169 negative 2922-83-0 Kynurenic Acid up 1.32 0.049 0.243 negative 492-27-3 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Cysteine and Methionine

Metabolites S-Adenosylhomocysteine up 1.15 0.021 0.213 positive 979-92-0 Histidine Metabolism Carnosine up 1.31 0.037 0.230 negative 305-84-0 Arginine and Proline

Metabolites 4-Oxoproline up 1.30 0.003 0.102 positive 4347-18-6 Tyrosine Metabolites Vanylglycol (MHPG) up 1.64 0.011 0.169 negative 67423-45-4 Tyramine up 1.06 0.028 0.216 positive 51-67-2 Thymidine down 1.49 0.028 0.216 positive 50-89-5 Uracil down 1.38 0.049 0.243 negative 66-22-8 Purine Metabolism GMP up 1.14 0.028 0.230 negative 85-32-5 Miscellaneous Hydroxybutyric acid up 1.35 0.002 0.055 negative 5094-24-6 Methylglutarylcarnitine up 2.42 0.005 0.134 positive 102673-95-0 Trimethylamine N-oxide up 2.65 0.021 0.213 positive 1184-78-7 2-Hydroxypyridine up 1.36 0.037 0.275 positive 142-08-5 N-Acetylneuraminic Acid up 1.23 0.037 0.230 negative 131-48-6

bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Table 2. Differentially expressed targeted metabolites from the PVC of APOE3 vs. E4 mice

Regulation Fold Detection CAS Metabolite p-value FDR in E4/4 Change Mode Number Fatty Acids gamma-Linolenic Acid down 1.19 0.003 0.068 positive 506-26-3 Palmitic Acid down 1.32 0.021 0.212 negative 57-10-3 10-Hydroxydecanoate up 2.22 0.015 0.195 positive 1679-53-4 Docosahexaenoic Acid down 2.03 0.021 0.212 negative 6217-54-5 Arachidonic acid down 1.45 0.021 0.212 negative 506-32-1 Oligosaccharides trisaccharide up 1.82 0.015 0.212 negative tetrasaccharide up 2.70 0.021 0.212 negative Vitamin and Vitamin

Derivatives Dehydroascorbic acid up 2.03 0.001 0.062 positive 490-83-5 Phylloquinone up 1.90 0.001 0.062 positive 84-80-0 Ascorbic acid up 105.84 0.005 0.192 negative 50-81-7 Tocopherol up 2.28 0.008 0.192 negative 59-02-9 Energy-Related

Metabolites Acetylcarnitine up 1.40 0.003 0.068 positive 3040-38-8 Carnitine up 1.31 0.011 0.154 positive 541-15-1 Coenzyme A (CoA) up 2.82 0.028 0.282 positive 85-61-0 Pyruvate up 1.43 0.028 0.238 negative 127-17-3 Acetoacetic Acid up 1.40 0.028 0.238 negative 541-50-4 Cholesterol Metabolites Taurodeoxycholate up 1.94 0.003 0.068 positive 516-50-7 Cholesteryl Acetate up 2.24 0.008 0.192 negative 604-35-3 Lanosterol up 2.50 0.011 0.192 negative 79-63-0 Amino Acids Tyrosine up 1.25 0.037 0.270 negative 60-18-4 Serine down 1.18 0.049 0.270 negative 56-45-1 Glutamate down 1.08 0.049 0.282 positive 56-86-0 Cysteine and Methionine

Metabolites Pterin down 3.91 0.002 0.192 negative 2236-60-4 Histidine Metabolites Urocanic acid down 1.34 0.021 0.212 negative 104-98-3 Carnosine up 1.32 0.028 0.282 positive 305-84-0 Arginine and Proline

Metabolites bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Phosphocreatine up 1.62 0.037 0.282 positive 67-07-2 4-Guanidinobutyric Acid down 1.09 0.049 0.270 negative 463-00-3 Threonine Metabolites 2-Ketobutyric Acid up 1.43 0.021 0.212 negative 600-18-0 Tyrosine Metabolites Homovanillic Acid down 1.30 0.049 0.270 negative 306-08-1 Riboflavin Metabolites 2,4-Dihydroxypteridine up 2.07 0.008 0.154 positive 487-21-8 (Lumazine) Flavin adenine dinucleotide up 1.19 0.021 0.212 negative 146-14-5 (FAD) Lumichrome down 1.71 0.049 0.282 positive 1086-80-2 Pyrimidine Metabolites 2-Aminoisobutyric acid up 1.38 0.011 0.154 positive 62-57-7 UMP up 1.22 0.049 0.270 negative 58-97-9 CMP up 1.29 0.049 0.282 positive 63-37-6 Miscellaneous Lipoic Acid up 3.76 0.005 0.192 negative 1077-28-7 Thiourea down 1.45 0.008 0.192 negative 62-56-6 Butyrylcarnitine up 1.24 0.011 0.154 positive 25576-40-6 Hydroxybutyric Acid up 1.49 0.011 0.192 negative Indoxyl Sulfate down 2.12 0.021 0.212 negative 2642-37-7 3-Hydroxymethylglutaric Acid up 1.27 0.028 0.238 negative 503-49-1 2-Hydroxypyridine up 1.45 0.037 0.282 positive 142-08-5 Maleic acid up 1.92 0.037 0.270 negative 110-16-7 N-Acetylneuraminic Acid up 1.17 0.037 0.282 positive 131-48-6 N-Methylglutamic Acid down 1.19 0.049 0.270 negative 35989-16-3 2-Aminoadipic Acid down 1.17 0.049 0.270 negative 542-32-5 Cytidine diphosphate choline down 1.17 0.049 0.282 positive 987-78-0 (CDPcholine) Glutathione up 1.49 0.049 0.282 positive 70-18-8

bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Table 3. GO biological processes observed in the PIUMet analysis of differentially expressed

untargeted metabolites from the EC of APOE4/4 vs. APOE3/3 mice

GO Term Enrichment p-value FDR

carboxylic acid transport 12.35 9.80E-20 1.43E-16 carboxylic acid transmembrane transport 22.34 1.36E-16 1.04E-13 long-chain fatty acid metabolic process 18.08 2.77E-13 1.50E-10 fatty acid metabolic process 7.8 1.45E-11 5.40E-09 indolalkylamine metabolic process 44.68 2.48E-09 6.68E-07 disaccharide metabolic process 63.29 6.92E-09 1.77E-06 IMP metabolic process 48.69 3.47E-08 7.43E-06 steroid metabolic process 6.87 1.78E-07 3.24E-05 maltose metabolic process 126.58 4.81E-07 8.06E-05 kynurenine metabolic process 50.63 7.51E-07 1.18E-04 IMP biosynthetic process 50.63 7.51E-07 1.19E-04 tryptophan catabolic process 50.63 7.51E-07 1.20E-04 'de novo' IMP biosynthetic process 63.29 9.46E-06 1.13E-03 ketone catabolic process 3.90E-05 3.92E-03 dopamine metabolic process 20.25 4.12E-05 4.06E-03 neurotransmitter catabolic process 37.97 5.54E-05 5.25E-03 keratan sulfate catabolic process 31.65 1.00E-04 8.37E-03 vitamin metabolic process 6.38 3.60E-04 2.35E-02

bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Table 4. Metabolite concentrations in the EC of APOE3/3 and APOE4/4 mice, as observed by 1H

NMR spectroscopy

Metabolite APOE3/3 APOE4/4 p-value concentration (mM) concentration (mM)

Creatine (Cr) 3.73 +/- 0.25 3.09 +/- 0.21 3.04E-04

Phosphocreatine (PCr) 0.84 +/- 0.27 2.55 +/- 0.18 1.00E-05

Gamma-aminobutyrate (GABA) 2.14 +/- 0.12 1.98 +/- 0.13 0.036

Glutamate (Glu) 7.65 +/- 0.15 6.58 +/- 0.17 5.30E-05

Glycerophosphocholine (GPC) 1.25 +/- 0.12 0.67 +/- 0.07 2.90E-05

Taurine (Tau) 6.16 +/- 0.14 5.31 +/- 0.15 1.59E-05

Cr+PCr 5.58 +/- 0.11 5.85 +/- 0.12 8.90E-04

Glu+Gln 10.43 +/- 0.21 9.22 +/- 0.23 2.30E-05

NAA+NAAG 6.19 +/-0.14 5.81 +/- 0.12 8.90E-04

bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Table 5. Differentially expressed targeted lipids from the EC of APOE4/4 vs. APOE3/3 mice

Regulation Fold Metabolite p-value in E4/4 Change Cholesteryl esters CE 18:0 down 1.44 0.039 Diacylglycerols DAG 28:0/14:0 down 1.33 0.003 DAG 30:0/14:0 down 1.50 0.002 DAG 32:0/16:0 down 1.30 0.006 DAG 32:1/16:0 down 1.45 0.001 DAG 32:2/16:1 down 1.38 0.031 DAG 34:1/16:0 down 1.28 0.011 DAG 34:2/16:0 down 1.38 0.016 DAG 38:3/18:0 down 1.24 0.032 DAG 38:4/18:1 down 1.32 0.022 DAG 40:6/18:0 down 1.19 0.013 Ceramides Cer 18:1 up 1.60 2.00E-04 Sphingomyelins SM 18:0 down 1.09 0.045 Hexosylceramides

(glucosyl- and galactosyl-) HexCer 16:0 up 2.45 1.19E-05 HexCer 16:1 up 1.50 0.001 HexCer 18:0 up 1.43 0.020 HexCer 20:0 up 1.47 0.027 HexCer 26:0 up 2.08 0.049 Sulfatides Sulf(2OH) 18:0 up 1.31 0.028 Lactosylceramides LacCer 16:0 up 1.81 0.005 Phosphatidic acids PA 40:6 up 1.18 0.044 Phosphatidylcholines PC 30:0 down 1.14 0.045 PC 36:0 up 1.11 0.006 PC 42:7 up 1.28 0.020 Phosphatidylinositols PI 38:3 down 1.39 0.003 PI 38:4 down 1.32 0.033 Phosphatidylserines bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

PS 42:4 up 1.33 0.022 PS 42:5 up 1.29 0.031 Lyso ether

phosphatidylcholines LPCe 18:0 up 1.14 0.010 Lyso phosphatidylinositols LPI 16:0 down 1.15 0.002 Bis(monoacylglycero)

phosphates, LBPA BMP 32:0 up 1.41 0.003 BMP 34:0 up 1.39 0.014 BMP 34:1 up 1.37 0.013 BMP 36:0 up 1.75 0.011 BMP 36:1 up 1.44 0.014 BMP 38:4 up 1.95 0.011

bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Table 6. Differentially expressed targeted lipids from the PVC of APOE4/4 vs. APOE3/3 mice

Regulation Fold Metabolite p-value in E4/4 Change Diacylglycerols DAG 38:3/18:1 down 1.47 0.023 DAG 40:5/18:0 down 1.41 0.014 Ceramides Cer 18:1 down 1.30 0.017 Sphingomyelins SM 16:0 down 1.30 0.015 SM 20:0 down 1.27 0.011 Dihydrosphingomyelins dhSM 16:0 down 1.64 0.020 Hexosylceramides

(glucosyl- and galactosyl-) HexCer 16:1 down 1.56 3.16E-04 Ether

phosphatidylcholines PCe 36:0 down 1.26 0.043 PCe 40:6 down 1.29 0.005 Lyso ether

phosphatidylcholines LPCe 16:0 down 1.53 0.007 Bis(monoacylglycero)

phosphates, LBPA BMP 32:0 up 1.66 0.041 BMP 32:1 up 1.54 0.031 BMP 34:0 up 1.62 0.035 BMP 34:2 up 1.36 0.007 N-acyl

phosphatidylserines NAPS 38:4 up 2.09 0.024

bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Supplementary Material

APOE4 is Associated with Differential Regional Vulnerability to Bioenergetic

Deficits in Aged APOE Mice

Tal Nuriel1,2,*, Delfina Larrea1,3, David N. Guilfoyle4, Leila Pirhaji5, Kathleen Shannon6, Hirra Arain1,2,

Archana Ashok1,2, Marta Pera1,3, Qiuying Chen7, Allissa A. Dillman8, Helen Y. Figueroa1,2, Mark R.

Cookson8, Steven S. Gross7, Ernest Fraenkel5,9, Karen E. Duff1,2,10, †, and Estela Area-Gomez1,3, †

* Corresponding author: Tal Nuriel, email: [email protected]

†These two authors contributed equally to this work and are co-senior authors.

bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

FIGURE LEGENDS

Supplementary Fig. S1. Seahorse analysis reveals decreased mitochondrial respiration in the PVC, but not in the EC, of aged APOE4 mice. Seahorse analysis was performed in order to analyze the effects of differential APOE isoform expression on mitochondrial respiration in mitochondria that were purified from the PVC and EC of aged APOE mice (4 APOE4/4 males, tissues pooled vs. 4 APOE3/3 males, tissues pooled). (A) The complex I- and complex II-driven ETC activity from each region shows decreased mitochondrial respiration in the mitochondria purified from the PVC, but not the EC of the aged APOE4/4 mice. (B-C) Bar graphs showing the average oxygen consumption rate (OCR) from (B) State 3 and for (C) the Respiration Control Ratio (RCR; state 3u/state 4o) in mitochondria purified from each region of the APOE4/4 mice, as a percentage of the APOE3/3 OCR from the equivalent tissues. The dotted blue line represents the normalized levels in the APOE3/3 tissues. (* denotes p < 0.05; ** denotes p < 0.01; *** denotes p < 0.001; **** denotes p < 0.0001)

Supplementary Fig. S2. Western blot and qPCR analysis shows no significant effects of APOE4 on mitochondrial mass or ETC protein expression. Western blotting and qPCR analysis was performed in order to investigate the effects of differential APOE isoform expression on various markers of mitochondrial mass and ETC proteins in the EC, Hip and Ctx of 21-month-old APOE4/4 vs. APOE3/3 mice. (A-B) Western blot analysis did not reveal any unique differences in the levels of mitochondrial outer membrane protein Tom20 or in ETC complex subunits between genotypes in the EC. Likewise, (C) the ratio of Mitochondrial:Nuclear DNA, as well as (D) the levels of PGC1α RNA were unchanged between APOE genotypes in the EC of 21-month-old APOE mice. (* denotes p < 0.05; ** denotes p < 0.01)

Supplementary Fig. S3. Results network from the PIUMet analysis of differentially expressed untargeted metabolites from the EC of APOE4/4 vs. APOE3/3 mice. Shown here is the results-network that was generated from the PIUMet (Prize-collecting Steiner forest algorithm for Integrative Analysis of Untargeted Metabolomics) analysis. PIUMet was able to predict identities for the differential expressed unidentified metabolite features or peaks detected in the EC of aged APOE4/4 vs. APOE3/3 mice (corrected p-value <0.05, 4 APOE4/4 males, tissues pooled vs. 4 APOE3/3 males, tissues pooled). The resulting network shows 242 nodes connected by 328 edges. Input metabolite peaks or features are shown as red triangles, while grey circles represent proteins and square nodes display the metabolites in the network. The metabolites directly connected to the differential metabolite features show their putative identities. The size of the nodes reflects the specificity score and the thickness of the edges the confidence of the interaction. In addition, the robustness score of each node is associated with the thickness of each node line width.

bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Supplementary Fig. S1

bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Supplementary Fig. S2

bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Supplementary Fig. S3 (high-res version available as supplemental file)

bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Supplementary Table S1. Differentially expressed genes from the EC of APOE3/4 vs. APOE3/3 mice

log2 Gene Name p-value FDR FoldChange

Serpina3m 3.97 1.86E-166 3.41E-162

Serpina3n 1.95 4.65E-147 4.26E-143

Oscar 2.20 1.09E-60 6.67E-57

Rdh13 -0.44 1.37E-24 6.29E-21

Myadm 0.34 3.96E-14 1.45E-10

Thnsl1 0.50 4.69E-13 1.43E-09

Gm12762 1.14 4.52E-12 1.18E-08

Mboat7 0.21 6.98E-12 1.60E-08

Ica1 -0.35 1.55E-11 2.57E-08

Syngr3 0.26 1.38E-11 2.57E-08

Tmc4 0.62 1.42E-11 2.57E-08

Asprv1 -0.71 1.36E-10 1.92E-07

Fbxl2 0.18 1.35E-10 1.92E-07

Gm1082 1.04 1.76E-10 2.30E-07

Gm1078 -0.61 5.46E-10 6.67E-07

Wdfy1 -0.52 6.90E-10 7.90E-07

Srcap -0.38 9.06E-10 9.63E-07

Zbtb20 -0.57 9.46E-10 9.63E-07

Dpysl2 -0.38 2.18E-09 2.00E-06

Pon2 -0.28 2.13E-09 2.00E-06

Gm15494 -0.59 3.57E-09 3.11E-06

Ccdc32 0.15 4.23E-09 3.37E-06

Mll2 -0.42 4.17E-09 3.37E-06 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Fam103a1 0.22 4.86E-09 3.71E-06

Pkd1 -0.23 5.58E-09 4.09E-06

Atp6v1e1 0.16 6.86E-09 4.83E-06

Zfp652 -0.22 7.28E-09 4.94E-06

Dchs1 -0.27 9.22E-09 6.03E-06

Gm4723 0.61 1.53E-08 9.68E-06

1520402A15Rik -0.52 1.85E-08 1.13E-05

Cml5 -0.74 2.63E-08 1.50E-05

Nfia -0.33 2.59E-08 1.50E-05

Atp6v0e2 0.14 2.85E-08 1.53E-05

Trcg1 -0.92 2.79E-08 1.53E-05

AC154449.2 -0.40 3.01E-08 1.57E-05

Zfhx3 -0.43 3.13E-08 1.59E-05

Atp6v1g2 0.16 6.11E-08 3.02E-05

Mios -0.31 6.36E-08 3.07E-05

Lrrc58 -0.37 8.70E-08 4.09E-05

Ppp6r2 -0.18 9.50E-08 4.24E-05

Trim72 -0.67 9.33E-08 4.24E-05

AC132303.1 -0.37 1.00E-07 4.27E-05

Cftr -0.71 9.84E-08 4.27E-05

Cybrd1 -0.59 1.05E-07 4.35E-05

Ptprh 0.80 1.39E-07 5.67E-05

Cdh10 0.35 1.43E-07 5.69E-05

AC102195.1 -0.88 1.51E-07 5.78E-05

Mapt 0.12 1.51E-07 5.78E-05

Gm10931 -0.71 1.55E-07 5.81E-05

Gm3435 -0.35 1.59E-07 5.82E-05 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

BX537302.1 -0.53 1.67E-07 5.87E-05

Itpr2 -0.21 1.66E-07 5.87E-05

Rgs4 0.32 1.70E-07 5.87E-05

Tgfa -0.23 1.93E-07 6.54E-05

Sbsn -0.44 2.00E-07 6.66E-05

Gm11944 -0.51 2.10E-07 6.87E-05

Prdm11 -0.16 2.27E-07 7.30E-05

Tmem170b -0.26 2.74E-07 8.65E-05

Plin4 -0.70 2.93E-07 9.09E-05

Ahsa1 0.17 3.39E-07 0.000102

Slco2a1 -0.70 3.37E-07 0.000102

Ksr2 -0.35 3.48E-07 0.000103

2010107E04Rik 0.17 3.65E-07 0.000106

Pan3 -0.24 4.46E-07 0.000128

RP24-161J8.2 -0.51 4.62E-07 0.00013

Nars 0.15 4.94E-07 0.000137

AC122417.1 -0.32 5.40E-07 0.000143

Spnb1 -0.29 5.24E-07 0.000143

Srp19 0.16 5.34E-07 0.000143

1500031L02Rik 0.18 6.00E-07 0.000151

Dctd -0.32 5.81E-07 0.000151

Gm12116 -0.76 5.94E-07 0.000151

Gm16997 -0.83 5.93E-07 0.000151

Adcy9 -0.25 6.33E-07 0.000157

Acacb -0.38 6.49E-07 0.000158

Igsf3 -0.36 6.54E-07 0.000158

Hif3a -0.72 6.81E-07 0.000162 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Syt11 0.15 7.71E-07 0.000181

Vopp1 0.26 8.24E-07 0.000191

Fmo2 -0.66 8.89E-07 0.000202

Ranbp17 -0.36 8.94E-07 0.000202

Srp54b 0.27 9.45E-07 0.000211

0610038L08Rik -0.55 9.59E-07 0.000212

Mrpl36 0.20 9.74E-07 0.000212

Aff1 -0.32 9.95E-07 0.000214

Rdh9 -0.58 1.03E-06 0.000217

Synj2 0.33 1.03E-06 0.000217

Stmn1 0.24 1.05E-06 0.000218

Hbxip 0.19 1.08E-06 0.000221

Zc3h7b 0.22 1.09E-06 0.000221

Anxa4 0.62 1.12E-06 0.000222

Ndufb6 0.17 1.13E-06 0.000222

Yeats4 0.18 1.12E-06 0.000222

Ngrn 0.15 1.26E-06 0.000246

Bid -0.28 1.35E-06 0.00026

Ncald 0.26 1.44E-06 0.000275

Flnc -0.55 1.48E-06 0.00028

CT868690.1 -0.79 1.53E-06 0.000286

2410004B18Rik 0.18 1.61E-06 0.000298

Atf4 0.26 1.80E-06 0.00033

Icosl -0.48 1.86E-06 0.000338

Tmem60 0.23 1.97E-06 0.00035

Zfp334 -0.13 1.97E-06 0.00035

Txndc12 0.18 2.08E-06 0.000367 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Dync1h1 -0.17 2.11E-06 0.000369

Sdc4 -0.39 2.39E-06 0.000412

Scube2 -0.47 2.45E-06 0.000419

Agl -0.15 2.48E-06 0.00042

Ahnak -0.46 2.53E-06 0.000425

Pkp2 -0.64 2.64E-06 0.000435

Tmem85 0.13 2.62E-06 0.000435

Snrpd3 0.18 2.69E-06 0.00044

Celsr1 -0.33 2.77E-06 0.00045

Ahnak2 -0.44 2.80E-06 0.00045

Gm13375 -0.41 2.87E-06 0.000458

Zfand6 0.20 2.96E-06 0.000467

Selk 0.15 3.07E-06 0.000481

Gpnmb -0.47 3.15E-06 0.000487

Gpr182 -0.50 3.17E-06 0.000487

Zhx3 -0.26 3.19E-06 0.000487

Rab7 0.10 3.32E-06 0.000503

Zbtb44 -0.32 3.38E-06 0.000507

RP23-388P16.8 -0.78 3.46E-06 0.000515

Ncbp2 0.13 3.54E-06 0.000523

Slc25a22 0.18 3.88E-06 0.000568

Cobl 0.27 3.93E-06 0.000572

Cd99l2 0.13 4.05E-06 0.000585

Ndfip1 0.20 4.44E-06 0.000626

Slc1a4 0.16 4.39E-06 0.000626

Trp73 -0.70 4.42E-06 0.000626

Ankrd7 -0.79 4.57E-06 0.00064 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Prdx2 0.10 4.62E-06 0.000641

Agxt2l2 -0.21 4.78E-06 0.000651

Ly6e 0.23 4.76E-06 0.000651

Pm20d1 -0.45 4.80E-06 0.000651

Dhdh -0.26 5.03E-06 0.000677

Gm15594 -0.73 5.33E-06 0.000712

Ppp1r16b -0.29 5.42E-06 0.000719

Med1 -0.17 5.67E-06 0.000733

Pak1 0.18 5.68E-06 0.000733

Uqcrb 0.22 5.68E-06 0.000733

Wnk1 -0.16 5.64E-06 0.000733

Gm10151 -0.62 5.85E-06 0.000749

Isca2 0.15 6.19E-06 0.000787

4631405J19Rik -0.62 6.80E-06 0.000859

Nlk 0.22 7.22E-06 0.000906

Gm10595 -0.73 7.72E-06 0.000962

Atp6v1b2 0.14 8.34E-06 0.000991

Atp6v1g1 0.17 8.39E-06 0.000991

Cldnd1 0.13 8.13E-06 0.000991

Cyb5 0.19 8.32E-06 0.000991

Dnajb6 0.14 8.05E-06 0.000991

Fxyd4 -0.61 8.41E-06 0.000991

Lonrf3 -0.36 8.50E-06 0.000991

Nicn1 0.14 8.40E-06 0.000991

Stx12 0.15 8.51E-06 0.000991

Tspyl4 0.18 8.55E-06 0.000991

Ube2e1 0.13 8.51E-06 0.000991 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

BC046401 -0.55 8.84E-06 0.001018

Rab28 0.14 8.90E-06 0.001019

Map1lc3b 0.19 9.11E-06 0.001037

Lefty1 -0.67 9.18E-06 0.001038

Ube2b 0.18 9.52E-06 0.00107

Lama1 -0.47 9.77E-06 0.001092

Sntb2 -0.33 9.87E-06 0.001096

Gm15082 -0.73 9.96E-06 0.001099

Cd2bp2 0.10 1.00E-05 0.0011

Mtap1a -0.16 1.02E-05 0.001114

Slc2a9 -0.57 1.03E-05 0.001118

Rexo2 0.14 1.06E-05 0.001143

Atcay 0.15 1.07E-05 0.001144

Otud1 0.25 1.07E-05 0.001144

Slc25a3 0.12 1.12E-05 0.001186

Arl6ip5 0.13 1.13E-05 0.001193

Lysmd4 0.13 1.17E-05 0.001222

Ywhah 0.14 1.17E-05 0.001222

Atp6ap1 0.11 1.24E-05 0.001274

Fnbp1l 0.26 1.24E-05 0.001274

Stab2 -0.36 1.24E-05 0.001274

Fbxl18 -0.20 1.28E-05 0.001298

Letmd1 0.12 1.38E-05 0.001389

Txnl4a 0.21 1.38E-05 0.001389

Lysmd2 0.17 1.40E-05 0.001405

Nup98 -0.17 1.43E-05 0.001423

Gm12394 -0.46 1.48E-05 0.001465 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

7SK.295 -0.70 1.49E-05 0.001469

Atg5 0.15 1.53E-05 0.001496

Ndfip2 0.23 1.58E-05 0.001541

Elmo3 -0.28 1.65E-05 0.001601

Stx8 0.19 1.71E-05 0.001644

AC159277.1 -0.47 1.74E-05 0.001654

Col6a6 -0.64 1.74E-05 0.001654

Trpm3 -0.42 1.74E-05 0.001654

SNORD38.1 -0.65 1.78E-05 0.001676

Prox1 -0.63 1.84E-05 0.001728

2610507B11Rik -0.11 1.95E-05 0.001805

Ndrg4 0.15 1.93E-05 0.001805

Rasgef1b 0.24 1.95E-05 0.001805

Notch2 -0.29 2.03E-05 0.001861

Snrnp27 0.15 2.02E-05 0.001861

Mrrf 0.14 2.05E-05 0.001869

Hipk2 -0.26 2.08E-05 0.001882

Lpp -0.35 2.09E-05 0.001883

Arpp19 0.26 2.20E-05 0.001957

Gm15893 -0.41 2.20E-05 0.001957

Rbm18 0.15 2.18E-05 0.001957

1700021F05Rik 0.16 2.29E-05 0.00201

Slc24a4 -0.30 2.29E-05 0.00201

Ubr4 -0.15 2.28E-05 0.00201

AI413582 0.19 2.38E-05 0.002073

Slc39a10 0.22 2.45E-05 0.00213

Spcs2 0.13 2.48E-05 0.002147 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

1110054O05Rik 0.13 2.54E-05 0.002151

Hsp90b1 0.20 2.53E-05 0.002151

Nup214 -0.16 2.50E-05 0.002151

Patz1 -0.18 2.53E-05 0.002151

Galntl2 -0.68 2.56E-05 0.002164

Zfp783 -0.25 2.73E-05 0.002293

Got1 0.16 2.87E-05 0.002403

Maml2 -0.30 2.89E-05 0.00241

Qrfpr 0.48 2.92E-05 0.002422

Cd200 0.23 2.97E-05 0.002448

Sra1 0.16 3.00E-05 0.002464

Cpne4 0.23 3.05E-05 0.002481

Serf1 0.23 3.04E-05 0.002481

Igsf10 -0.35 3.15E-05 0.002547

Ralyl 0.21 3.17E-05 0.002547

Tsku -0.48 3.17E-05 0.002547

Foxo3 -0.24 3.22E-05 0.002577

Crebbp -0.21 3.24E-05 0.002579

Pnpla7 -0.28 3.32E-05 0.002622

Ykt6 0.12 3.32E-05 0.002622

B230120H23Rik -0.35 3.35E-05 0.002637

Lrrc29 -0.41 3.44E-05 0.002682

Rgs20 0.23 3.44E-05 0.002682

Klhl11 -0.29 3.48E-05 0.002688

Ptprn2 0.23 3.47E-05 0.002688

Cisd2 0.14 3.57E-05 0.002734

Pacsin2 0.22 3.57E-05 0.002734 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Gm15609 -0.41 3.75E-05 0.002862

Hnrnph2 0.17 3.86E-05 0.00292

Nf1 -0.21 3.84E-05 0.00292

Nap1l3 0.26 3.94E-05 0.002969

AL662835.2 -0.63 3.99E-05 0.002981

Tmem14c 0.18 3.99E-05 0.002981

Ube2e2 0.13 4.12E-05 0.003067

Gm6588 -0.56 4.17E-05 0.003092

4933425O20Rik -0.34 4.32E-05 0.003178

R3hdm2 0.19 4.34E-05 0.003178

Ttc28 -0.37 4.33E-05 0.003178

Npepps 0.12 4.40E-05 0.00321

Gm12576 -0.64 4.47E-05 0.003246

Ssbp1 0.15 4.48E-05 0.003246

Kctd21 0.18 4.56E-05 0.003268

Tk1 -0.40 4.56E-05 0.003268

Tmem9b 0.13 4.57E-05 0.003268

Pitpna 0.14 4.59E-05 0.00327

Crb1 0.58 4.66E-05 0.003311

Nt5dc2 -0.39 4.74E-05 0.003326

Pfdn1 0.14 4.73E-05 0.003326

Vdac1 0.10 4.73E-05 0.003326

Ostc 0.17 4.76E-05 0.003329

Cmpk1 0.20 4.79E-05 0.003339

Mapre2 0.09 4.86E-05 0.003369

Apbb2 -0.14 5.01E-05 0.00346

Vta1 0.13 5.05E-05 0.003476 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Pfkp 0.15 5.22E-05 0.003581

2210020M01Rik -0.47 5.24E-05 0.003585

Tmub2 0.14 5.26E-05 0.003585

Tet2 -0.20 5.36E-05 0.003629

Ythdf1 0.11 5.37E-05 0.003629

Spdef -0.62 5.43E-05 0.003659

Slc14a1 -0.30 5.53E-05 0.003705

Timm22 0.14 5.54E-05 0.003705

Golm1 -0.24 5.66E-05 0.00377

A930024N18Rik -0.41 5.70E-05 0.003771

Muc6 -0.52 5.70E-05 0.003771

Siglec1 -0.45 5.81E-05 0.003831

Morf4l1 0.15 5.97E-05 0.003889

Proca1 -0.24 5.94E-05 0.003889

Ror1 -0.58 5.96E-05 0.003889

Lifr -0.27 6.01E-05 0.003908

AC125486.1 -0.54 6.30E-05 0.00401

B3galnt1 0.19 6.30E-05 0.00401

Pde6h -0.46 6.27E-05 0.00401

Polh -0.23 6.26E-05 0.00401

Tnfrsf19 0.20 6.27E-05 0.00401

Tspan18 -0.56 6.21E-05 0.00401

Mff 0.12 6.33E-05 0.004016

Grinl1a 0.11 6.43E-05 0.004061

Tulp3 -0.20 6.49E-05 0.004083

Leprotl1 0.14 6.62E-05 0.004153

Gm11721 -0.47 6.67E-05 0.00417 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Mkks 0.21 6.76E-05 0.004211

Plec -0.24 6.98E-05 0.00432

Rbx1 0.11 6.98E-05 0.00432

Tgfbr3 -0.27 7.09E-05 0.004376

Nt5c1a -0.33 7.19E-05 0.004391

Slc1a2 -0.19 7.17E-05 0.004391

Tbcd -0.10 7.19E-05 0.004391

Calm2 0.16 7.23E-05 0.004398

Exd2 0.18 7.42E-05 0.004502

Lca5l -0.32 7.51E-05 0.004524

Ywhab 0.12 7.49E-05 0.004524

6330527O06Rik 0.36 7.55E-05 0.004533

D430018E03Rik -0.42 7.62E-05 0.004533

Plce1 -0.43 7.62E-05 0.004533

Vps24 0.11 7.60E-05 0.004533

BC031181 0.11 7.69E-05 0.004547

P2ry1 -0.49 7.72E-05 0.004547

Sema5a -0.28 7.70E-05 0.004547

Sdccag8 0.15 7.75E-05 0.004552

A330049M08Rik -0.55 7.96E-05 0.004662

Zfp365 0.17 8.03E-05 0.004687

Slmo2 0.12 8.14E-05 0.004734

Napg 0.14 8.19E-05 0.004744

Tor3a -0.16 8.21E-05 0.004744

Ace -0.27 8.36E-05 0.004816

Rec8 -0.63 8.50E-05 0.004881

Ppp2cb 0.11 8.53E-05 0.004886 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Rc3h2 -0.29 8.71E-05 0.00497

Ppfia1 -0.15 8.82E-05 0.00502

Cyp3a13 -0.64 8.87E-05 0.00503

Rbm15 -0.25 8.97E-05 0.005073

Stk32a -0.46 9.11E-05 0.005136

Bcs1l 0.15 9.20E-05 0.00517

4833412C05Rik 0.60 9.40E-05 0.005172

B430010I23Rik -0.67 9.37E-05 0.005172

Dlx1 0.19 9.24E-05 0.005172

Gm16201 -0.67 9.29E-05 0.005172

Ict1 0.16 9.38E-05 0.005172

P2ry13 0.22 9.32E-05 0.005172

Ppp2r5c 0.13 9.31E-05 0.005172

Fkbp5 -0.42 9.44E-05 0.00518

Ccdc6 -0.23 9.58E-05 0.005191

Chac2 0.26 9.53E-05 0.005191

Hat1 0.23 9.51E-05 0.005191

Pkd2l2 0.16 9.55E-05 0.005191

Ube2j2 0.11 9.68E-05 0.005229

1700121F15Rik -0.36 9.74E-05 0.005246

1500003O03Rik 0.11 9.99E-05 0.005365

Dctn3 0.15 0.0001 0.005377

Ccdc72 0.19 0.000101 0.00539

Trappc2 0.23 0.000101 0.00539

Eif1 0.12 0.000102 0.005403

Slc35f4 0.17 0.000102 0.005403

Bet1 0.24 0.000103 0.005433 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Gm15478 -0.52 0.000103 0.005433

4933401P06Rik -0.55 0.000105 0.005494

Eif2ak1 0.11 0.000105 0.005494

Ssh1 -0.14 0.000105 0.005494

Ttc13 -0.12 0.000106 0.005523

Arpc3 0.14 0.000107 0.005529

4831426I19Rik -0.43 0.000109 0.005615

Agbl2 -0.55 0.000109 0.005615

Zbtb1 -0.15 0.000109 0.005619

Grin2a -0.33 0.00011 0.005625

4921513D23Rik -0.18 0.000111 0.005651

Hcfc1r1 0.18 0.000111 0.005651

Aldh3a1 -0.49 0.000112 0.005691

Fnip2 -0.22 0.000112 0.005691

Man2a2 -0.12 0.000114 0.005748

Dnahc9 -0.51 0.000115 0.005829

Naip1 -0.50 0.000117 0.005864

Fam168b 0.07 0.000118 0.005893

Zfp398 -0.17 0.000118 0.005893

Elk4 -0.15 0.000119 0.005961

Gm16183 -0.32 0.000121 0.006002

Prkacb 0.18 0.000122 0.006039

Atp6v0d1 0.13 0.000123 0.006049

Gm15835 -0.46 0.000123 0.006049

Morn1 -0.25 0.000122 0.006049

Slc2a4 -0.49 0.000123 0.006049

Fat1 -0.38 0.000125 0.006118 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Tnxb -0.47 0.000125 0.006118

1500011B03Rik 0.14 0.000126 0.006119

Gm16982 -0.40 0.000127 0.006146

Rabgef1 0.15 0.000127 0.006146

Lmo4 0.28 0.000128 0.006206

Amn1 0.19 0.000129 0.006233

AC123699.1 -0.29 0.00013 0.00626

Agt -0.37 0.000131 0.006274

Gm12592 -0.26 0.000132 0.006274

Gm15892 -0.51 0.000132 0.006274

Svil -0.23 0.000132 0.006274

Gm15651 -0.62 0.000135 0.006382

Gm16854 -0.62 0.000135 0.006382

Rpl7l1 0.11 0.000136 0.006402

Trp53inp2 0.13 0.000137 0.006473

Tulp4 -0.16 0.000138 0.006473

Ero1lb 0.18 0.000138 0.006486

Gm16626 -0.46 0.000139 0.006487

Ube2f 0.12 0.000142 0.006602

Gm15869 -0.61 0.000142 0.006607

Gm16534 -0.59 0.000144 0.006687

Arhgap25 0.53 0.000145 0.006688

Pam16 0.14 0.000145 0.00669

Gm16334 -0.59 0.000146 0.006733

B230319C09Rik -0.43 0.000148 0.006781

1110007A13Rik -0.16 0.000153 0.007

Gde1 0.14 0.000154 0.007028 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Arhgef12 -0.12 0.000157 0.007135

Rnft2 0.14 0.000157 0.007137

Fam133b 0.13 0.000157 0.007139

Shisa6 -0.53 0.000158 0.007144

AA986860 -0.30 0.00016 0.007201

Cyp2d22 -0.18 0.000161 0.007231

Fanci -0.38 0.000161 0.007231

4930404I05Rik -0.35 0.000163 0.007302

Sept8 0.16 0.000163 0.007302

Dok5 0.24 0.000164 0.00732

5830444B04Rik -0.53 0.000165 0.007341

A2m -0.62 0.000166 0.007342

Sema3a 0.36 0.000167 0.007385

Golgb1 -0.11 0.000168 0.00742

Oplah -0.22 0.000169 0.007422

C030046E11Rik -0.13 0.00017 0.007474

AC122296.1 -0.58 0.000171 0.007475

Cyp2u1 -0.20 0.000172 0.00752

Dap3 0.10 0.000177 0.007682

Dnajc8 0.11 0.000177 0.007682

Gng11 0.37 0.000177 0.007682

Nsd1 -0.16 0.000177 0.007682

Rad54l -0.40 0.000178 0.007682

AC163296.1 -0.41 0.00018 0.007777

Fam151b 0.24 0.000181 0.007802

Tmem65 0.20 0.000182 0.007823

Fbxw7 0.18 0.000187 0.007968 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Gm14286 -0.40 0.000187 0.007968

Mdh1 0.15 0.000187 0.007968

Rab6 0.21 0.000187 0.007968

Nova2 -0.19 0.000188 0.007988

Slco1c1 0.18 0.000191 0.008098

Ppp2r2a 0.14 0.000193 0.008138

Dnaja1 0.18 0.000196 0.008172

F630040L22Rik -0.54 0.000196 0.008172

Gm12216 -0.53 0.000196 0.008172

Nav2 -0.18 0.000196 0.008172

Nkiras1 0.14 0.000195 0.008172

Stmn2 0.14 0.0002 0.008326

BC060267 -0.29 0.000201 0.008337

Fscn2 -0.49 0.000201 0.008337

6530402F18Rik -0.24 0.000202 0.008357

Dnaja3 0.09 0.000203 0.008394

Ehmt1 -0.12 0.000205 0.008443

Psma5 0.13 0.000206 0.008443

Mdk -0.26 0.000209 0.008535

Tshb -0.62 0.000209 0.008535

Ubl4 0.11 0.000209 0.008535

Zxdc -0.12 0.000211 0.008597

C230096C10Rik -0.13 0.000213 0.008627

Cpxm1 -0.28 0.000213 0.008627

Snap25 0.14 0.000213 0.008627

Idh3a 0.14 0.000214 0.008655

Cckbr 0.29 0.000215 0.008657 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

8030462N17Rik -0.18 0.000217 0.008737

Sat1 0.23 0.000219 0.008774

Chst2 -0.22 0.000222 0.008899

AC148089.1 -0.53 0.000223 0.008906

Pde4b 0.15 0.000224 0.008906

2610001J05Rik 0.15 0.000227 0.008983

Als2cr4 0.15 0.000228 0.008983

Fndc4 0.12 0.000227 0.008983

Gcnt7 -0.43 0.000227 0.008983

Tead1 -0.23 0.000229 0.009022

Chd4 -0.14 0.000232 0.00906

Eef2k -0.16 0.000232 0.00906

Kctd20 0.09 0.000231 0.00906

Ppp2r5a 0.13 0.000232 0.00906

Gpr161 -0.56 0.000232 0.009061

Bcas2 0.10 0.000233 0.00907

7SK.49 -0.48 0.000239 0.009277

Rbm47 -0.52 0.00024 0.009302

Tnfaip8l1 0.21 0.000244 0.009411

Wdr62 -0.30 0.000245 0.00943

Gm11827 -0.63 0.000245 0.009432

Uchl5 0.19 0.000247 0.009501

Gm16575 -0.61 0.000251 0.009616

Prkar2b 0.19 0.000252 0.009625

9830001H06Rik -0.40 0.000257 0.009775

Gm711 -0.46 0.000257 0.009775

Zfyve26 -0.12 0.000257 0.009775 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Itga11 -0.37 0.000259 0.009832

Acvr2b -0.27 0.000261 0.009887

Mpeg1 0.17 0.000262 0.009892

Btrc 0.09 0.000265 0.009896

Ncrna00081 0.15 0.000264 0.009896

Nutf2 0.13 0.000265 0.009896

Prr5 -0.42 0.000264 0.009896

Sardh -0.22 0.000263 0.009896

Zan -0.47 0.000265 0.009896

Cyb5r4 0.18 0.00027 0.010057

Invs -0.12 0.000271 0.010057

Ntrk3 -0.20 0.000271 0.010057

Actr1a 0.10 0.000275 0.010173

Atg12 0.14 0.000281 0.010392

Gm16896 -0.44 0.000284 0.010464

Gm14372 -0.47 0.000286 0.010492

Nsmce2 0.14 0.000286 0.010492

Slc16a12 -0.36 0.000286 0.010492

Gm15774 -0.32 0.000288 0.010501

Rab14 0.15 0.000288 0.010501

Trrap -0.14 0.000288 0.010501

Dctn5 0.10 0.000289 0.010517

Sipa1l3 -0.24 0.000292 0.01058

Hsp90ab1 0.10 0.000293 0.010603

Mrps14 0.15 0.000294 0.010625

Pole -0.39 0.000295 0.010625

Rab1 0.13 0.0003 0.010789 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Csnk1e 0.20 0.000301 0.010799

Rabif 0.12 0.000302 0.010831

Tbck -0.23 0.000304 0.010876

Vti1a 0.09 0.000305 0.010878

AC131761.1 -0.30 0.000312 0.011082

C130071C03Rik 0.19 0.000311 0.011082

9130011J15Rik 0.10 0.000312 0.011095

Yaf2 0.14 0.000314 0.011142

Mrps36 0.22 0.000316 0.011189

Il12a 0.55 0.00032 0.01128

D4Wsu53e 0.13 0.000321 0.011301

Iars 0.10 0.000321 0.011301

Pamr1 0.36 0.000327 0.011437

Pcmt1 0.12 0.000326 0.011437

Bag2 0.17 0.000328 0.01147

1700084C01Rik 0.24 0.000329 0.011471

N28178 0.17 0.000329 0.011471

Nap1l1 0.15 0.000331 0.011514

Pla1a -0.41 0.000333 0.011569

Megf6 -0.36 0.000336 0.011622

AC137156.2 -0.49 0.000337 0.011652

Lmln -0.15 0.000338 0.011673

Baalc 0.13 0.00034 0.011722

Syt13 0.15 0.000342 0.011765

Nynrin -0.25 0.000344 0.011812

Ndufc1 0.13 0.000346 0.01185

Klhdc8a 0.31 0.000348 0.011909 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Gm12227 -0.54 0.000349 0.011914

Atxn1 -0.21 0.000354 0.011998

Chordc1 0.23 0.000354 0.011998

Ndufa3 0.17 0.000354 0.011998

Pcp4 0.28 0.000355 0.011998

Tomm34 0.12 0.000355 0.011998

Gpr88 0.34 0.00036 0.0121

Myo7a -0.16 0.000359 0.0121

Nif3l1 0.12 0.000359 0.0121

Dnaja2 0.14 0.000371 0.012432

Lyst -0.23 0.000371 0.012438

Svep1 -0.59 0.000373 0.01248

Fam81a 0.18 0.000374 0.012489

Fbxl3 0.15 0.000375 0.012489

Cap2 0.18 0.000382 0.012658

Gm16316 -0.52 0.000382 0.012658

Igf2bp3 -0.28 0.000382 0.012658

Gm16738 -0.32 0.000383 0.01267

Cacng3 0.20 0.000388 0.012807

BC067074 -0.42 0.000391 0.012869

Pram1 -0.30 0.000394 0.012945

2900011O08Rik 0.10 0.000397 0.013025

Col18a1 -0.54 0.000397 0.013026

Atp6v1c1 0.13 0.000401 0.013028

Cfdp1 0.13 0.0004 0.013028

Dusp27 -0.52 0.0004 0.013028

Eif2b1 0.10 0.0004 0.013028 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Sdhd 0.09 0.000401 0.013028

AC121121.1 -0.27 0.000404 0.013051

Ankrd61 -0.32 0.000404 0.013051

Bnip3l 0.14 0.000404 0.013051

Rab2b 0.13 0.000405 0.013072

Cnnm4 -0.20 0.000409 0.013168

5730403B10Rik 0.08 0.00041 0.013183

Mef2c 0.28 0.000411 0.013202

Bbs1 0.10 0.000414 0.013258

Bhlhe41 -0.22 0.000416 0.013293

Kcng2 -0.55 0.000418 0.013347

AC117232.2 -0.36 0.000421 0.01335

Cct2 0.09 0.000421 0.01335

Drd5 -0.60 0.000421 0.01335

Tpcn2 -0.38 0.000419 0.01335

B3galt5 -0.35 0.000422 0.013361

Cuta 0.20 0.000423 0.013361

Hdac9 0.23 0.000424 0.013383

Rasl11b 0.29 0.000425 0.013383

Mrpl21 0.15 0.000426 0.013384

5730455O13Rik 0.21 0.000431 0.013526

Dnahc17 -0.32 0.000432 0.013543

Khdrbs3 0.21 0.000433 0.013547

Tomm22 0.11 0.000435 0.01357

Capn11 0.48 0.000438 0.013575

Cyp4f18 -0.49 0.000437 0.013575

Mtap1b -0.21 0.000436 0.013575 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Ndufc2 0.12 0.000438 0.013575

Paip2 0.13 0.000439 0.013575

Ap3s1 0.21 0.00044 0.013583

Trp53bp2 -0.18 0.000441 0.013608

Dnmt3a -0.14 0.000442 0.013613

Aloxe3 -0.32 0.000453 0.01392

Nudt4 0.20 0.000454 0.01392

Obscn -0.38 0.000455 0.013946

Ctgf 0.39 0.000457 0.013972

AA408865 -0.37 0.000461 0.014088

Pgk1 0.13 0.000466 0.01421

Gm1698 -0.60 0.000471 0.014346

Tmem52 -0.59 0.000475 0.014435

4933423P22Rik -0.54 0.000481 0.014582

Unc50 0.11 0.000483 0.014624

Ctbp2 -0.34 0.000487 0.014687

Rint1 -0.13 0.000487 0.014687

Znrf3 -0.23 0.000488 0.014703

Sec62 0.13 0.00049 0.014753

Gm11202 -0.53 0.000492 0.014762

Gm15965 -0.57 0.000493 0.014768

Mrpl42 0.18 0.000493 0.014768

C78339 0.19 0.000497 0.014794

Inpp4a -0.13 0.000496 0.014794

Mrps27 0.12 0.000496 0.014794

Tdgf1 -0.58 0.000501 0.014898

Airn -0.53 0.000502 0.014902 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Lrp1 -0.17 0.000505 0.014956

Sfxn3 0.17 0.000508 0.015024

5830432E09Rik -0.57 0.000512 0.015125

Atg4b 0.07 0.000513 0.015127

Pdcd5 0.13 0.000514 0.015127

Tmem49 0.14 0.000514 0.015127

Oxct1 0.15 0.000518 0.015198

6720456B07Rik 0.12 0.000521 0.015223

Klhl22 0.07 0.000521 0.015223

Sdhaf2 0.10 0.00052 0.015223

Cab39 0.13 0.000525 0.015315

Tert -0.29 0.000526 0.015323

Itgb7 -0.54 0.000529 0.015371

Naa38 -0.17 0.000532 0.015449

Msh4 -0.25 0.000533 0.015451

Slc36a2 -0.51 0.000536 0.015503

Lrrtm3 0.18 0.00054 0.01559

Cnot8 0.08 0.000544 0.015688

Atp5g1 0.12 0.000545 0.015691

2310003L22Rik 0.12 0.000549 0.015724

9230117E06Rik -0.57 0.00055 0.015724

Arfip2 0.07 0.000552 0.015724

Mt2 -0.33 0.000548 0.015724

Ndel1 0.11 0.000553 0.015724

Nphs2 -0.48 0.000549 0.015724

Pef1 0.15 0.000552 0.015724

Tnrc6c -0.12 0.000549 0.015724 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Gm10561 0.25 0.000554 0.015726

Slain1 0.12 0.000555 0.015727

Gm13110 -0.30 0.000556 0.015731

Dpy30 0.14 0.000561 0.01585

Gm14471 -0.47 0.000566 0.015956

Tdo2 -0.59 0.000566 0.015956

Arhgap21 -0.13 0.00057 0.015962

Farsb 0.11 0.00057 0.015962

Scamp5 0.09 0.00057 0.015962

Tbc1d4 -0.40 0.000569 0.015962

Mir494 -0.58 0.000572 0.015983

Trpv4 -0.35 0.000572 0.015983

Ankrd55 -0.33 0.000576 0.01603

Rabl2 0.15 0.000577 0.01603

Rtn4 0.18 0.000576 0.01603

Mpped2 0.23 0.000581 0.016119

Lamb3 -0.47 0.000584 0.016183

Ocel1 0.14 0.000586 0.016211

Fxyd7 0.32 0.000589 0.016275

Tceal8 0.19 0.000598 0.016501

Lrrc8a -0.21 0.000602 0.016585

4930521C21Rik -0.38 0.000608 0.016701

St18 -0.57 0.000609 0.016701

Ucp2 -0.32 0.000608 0.016701

Fbxo5 0.40 0.000616 0.016864

Cpne8 0.21 0.000624 0.017019

Ier3ip1 0.16 0.000624 0.017019 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Stx1a 0.25 0.000623 0.017019

Ctnna3 -0.53 0.000626 0.017031

Cml2 0.49 0.000632 0.01712

Ebf3 -0.52 0.00063 0.01712

Mkrn1 0.09 0.000633 0.01712

Rala 0.14 0.000631 0.01712

Alpk1 -0.24 0.000635 0.017123

Spen -0.16 0.000635 0.017123

Chst10 0.14 0.000638 0.017159

Igf2r -0.13 0.000639 0.017159

Trim62 -0.39 0.000638 0.017159

Enkur 0.35 0.000645 0.017288

RP23-145O4.5 -0.58 0.000649 0.017375

Shmt1 -0.28 0.00065 0.017375

Tspan4 -0.29 0.000651 0.017375

Sh3gl2 0.20 0.000654 0.017449

2810025M15Rik 0.20 0.000659 0.017547

Ada -0.38 0.000671 0.017781

Dlg1 0.11 0.00067 0.017781

Gm10564 -0.40 0.000669 0.017781

Il34 0.20 0.000675 0.017861

Wnt1 -0.37 0.000678 0.01793

Dnmbp -0.19 0.000684 0.018049

Ppara -0.29 0.000686 0.018091

Rufy3 0.11 0.000688 0.018111

Slc12a3 -0.54 0.000692 0.018202

Pion -0.20 0.000695 0.018229 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Tmem201 -0.17 0.000698 0.018295

Grb14 0.26 0.000702 0.01836

Fam164a 0.19 0.000705 0.018394

Fam192a 0.09 0.000705 0.018394

Tpm3 0.11 0.000714 0.018612

Aatf -0.13 0.000717 0.018655

SNORA17.403 -0.58 0.000718 0.018655

Crtc3 -0.16 0.000721 0.01872

Fbxl7 -0.42 0.000735 0.019037

4921522P10Rik -0.50 0.000738 0.019062

Arhgap31 -0.19 0.000738 0.019062

1110049F12Rik 0.11 0.000743 0.019143

Gm15816 -0.42 0.000742 0.019143

Ilf2 0.09 0.00075 0.019283

Stag3 -0.36 0.00075 0.019283

Rnf11 0.18 0.000755 0.019359

Txn1 0.14 0.00076 0.01948

Gm11750 -0.58 0.000762 0.019484

Lrrc20 0.11 0.000768 0.019583

Npas3 -0.22 0.000767 0.019583

Zc3h15 0.16 0.000769 0.019586

Heg1 -0.20 0.000771 0.019607

Cd27 -0.51 0.000776 0.019713

Map3k1 -0.28 0.000781 0.019793

Serinc3 0.15 0.000781 0.019793

Slc9a2 -0.44 0.000783 0.019803

Angptl7 -0.47 0.000785 0.01983 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Angptl2 -0.41 0.000791 0.019943

Fam159b 0.39 0.000791 0.019943

Casc1 -0.44 0.000798 0.020073

Rtn3 0.14 0.000801 0.020137

Reep5 0.13 0.000804 0.020173

Gm15800 -0.12 0.000807 0.0202

Gm15813 -0.40 0.000807 0.0202

1700112E06Rik 0.57 0.000812 0.020283

Ostm1 0.12 0.000815 0.020339

Tusc3 0.14 0.00082 0.020427

2810006K23Rik 0.19 0.000822 0.020452

Slc22a12 -0.52 0.000823 0.020466

Ttl 0.14 0.000825 0.020486

Loxhd1 -0.55 0.000833 0.020622

Tmem2 -0.20 0.000832 0.020622

Klf6 0.27 0.000837 0.020671

Tmem55a 0.11 0.000836 0.020671

Fam40b -0.44 0.000844 0.020786

Pappa -0.57 0.000844 0.020786

Acr -0.35 0.000846 0.020795

Ncam1 -0.15 0.000847 0.020795

3830431G21Rik -0.54 0.000859 0.021037

M6pr 0.10 0.000858 0.021037

Slitrk6 -0.53 0.000868 0.021223

Pvrl4 -0.35 0.000871 0.021246

RP23-389D15.1 -0.52 0.00087 0.021246

Vapa 0.14 0.000874 0.021256 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Zbtb42 -0.48 0.000873 0.021256

Mrpl9 0.10 0.000877 0.021317

Ccdc56 0.15 0.000887 0.0215

Fryl -0.15 0.000887 0.0215

Rab15 0.15 0.00089 0.021542

Bex1 0.19 0.000893 0.021591

Dgkg -0.28 0.000899 0.021702

Hnrnpa2b1 0.13 0.000901 0.021725

Ksr1 -0.31 0.000906 0.021811

Timp1 -0.36 0.000911 0.021903

Zcrb1 0.14 0.000914 0.021936

Bclp2 -0.56 0.000918 0.021972

Muc19 -0.42 0.000917 0.021972

Nlgn3 -0.17 0.00092 0.021972

Slc4a4 -0.19 0.00092 0.021972

Spats2l 0.17 0.000923 0.022022

Ryr1 -0.33 0.000926 0.022043

Tuba1a 0.15 0.000926 0.022043

Fank1 -0.24 0.000928 0.022048

Dlst 0.08 0.000929 0.022053

Rnf14 0.15 0.000932 0.022077

AC160104.2 -0.25 0.000941 0.022278

Foxm1 -0.32 0.000946 0.022342

Ubap1 0.11 0.000946 0.022342

Adamts18 -0.56 0.000952 0.022365

Ap3m2 0.10 0.000957 0.022365

Ep400 -0.09 0.000949 0.022365 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Fxyd6 0.18 0.000956 0.022365

Gmps -0.23 0.000953 0.022365

Lnx2 -0.31 0.000957 0.022365

Sema4b -0.23 0.000952 0.022365

Zfp568 -0.30 0.000956 0.022365

Lrp4 -0.23 0.00097 0.022636

Hnrnpul2 0.08 0.000972 0.022645

Timm23 0.09 0.000973 0.022645

4933431E20Rik 0.13 0.000976 0.022693

Pdhb 0.09 0.000988 0.022931

Tmem44 0.30 0.000989 0.022931

C4b -0.30 0.000996 0.023034

Gm14252 -0.56 0.000996 0.023034

Parva -0.15 0.000998 0.023056

Ap2m1 0.12 0.001 0.023062

Stmn4 0.18 0.001003 0.023103

Spcs1 0.11 0.001017 0.023418

Gm16154 -0.49 0.00103 0.023677

A430105I19Rik -0.25 0.001039 0.023842

Enoph1 0.11 0.001041 0.023865

Tmf1 -0.16 0.001045 0.023923

Cdc5l 0.11 0.001048 0.023975

Arsa 0.10 0.001052 0.024037

3110070M22Rik -0.43 0.001063 0.024182

Gm11186 -0.56 0.001063 0.024182

Oat 0.09 0.001062 0.024182

Llph 0.16 0.001065 0.024214 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

6330577E15Rik 0.14 0.001069 0.024258

Odf2 -0.11 0.001072 0.024294

4930578G10Rik -0.40 0.001081 0.024475

Batf3 -0.47 0.001083 0.024475

Cldn12 0.15 0.001083 0.024475

Atic 0.12 0.001086 0.024478

Thy1 0.12 0.001086 0.024478

4933427I04Rik -0.20 0.001091 0.024523

A830093I24Rik -0.25 0.001092 0.024523

Fahd1 0.15 0.001092 0.024523

AW011738 -0.26 0.001096 0.024537

Cacna1h -0.27 0.001099 0.024537

Gm14376 -0.55 0.001099 0.024537

Mobkl2b -0.32 0.001097 0.024537

Txnrd3 -0.29 0.0011 0.024537

Gm16197 -0.52 0.001104 0.024615

Basp1 0.16 0.001107 0.024647

Cyp4v3 -0.19 0.001123 0.024885

Gm14397 -0.56 0.001122 0.024885

Gm16907 -0.32 0.001122 0.024885

Usp29 -0.21 0.001123 0.024885

Snrpa 0.13 0.001129 0.02497

Tram1l1 0.12 0.001133 0.025008

Yars 0.10 0.001132 0.025008

Fam115a 0.13 0.001136 0.025042

Sdc1 -0.53 0.001137 0.025042

1300010F03Rik -0.13 0.001144 0.025162 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Rgag1 -0.29 0.001147 0.025193

Ivns1abp 0.14 0.001148 0.025195

Bsn -0.19 0.001152 0.025243

Gm15863 -0.38 0.001153 0.025243

Gm15265 -0.29 0.00116 0.025353

Capns2 -0.38 0.001168 0.025501

Sybu 0.09 0.001169 0.025501

Gng5 0.13 0.001178 0.025663

Adam22 -0.16 0.00118 0.025664

Pdia3 0.15 0.001194 0.025944

Kif3a 0.10 0.0012 0.026007

Map6d1 0.15 0.001199 0.026007

Amica1 -0.55 0.001207 0.026051

Ptprk 0.19 0.001204 0.026051

Scarna9 -0.48 0.001206 0.026051

Xdh -0.31 0.001206 0.026051

Dnase2a -0.19 0.001218 0.026137

Hrh1 0.23 0.001217 0.026137

Ints3 0.09 0.001216 0.026137

Mir486 -0.55 0.001217 0.026137

Telo2 -0.16 0.001218 0.026137

Golga7 0.11 0.001225 0.026247

Rnf180 0.17 0.001228 0.02628

Abhd6 0.14 0.001232 0.026338

Ensa 0.11 0.001238 0.026429

Kdm2b -0.15 0.001244 0.026536

Fry -0.13 0.001248 0.026568 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Gm7292 -0.16 0.001249 0.026568

Kdm5b -0.11 0.001261 0.026792

4930594M22Rik -0.45 0.001263 0.026802

Gm11681 -0.43 0.001265 0.026812

Gm10390 -0.52 0.001271 0.026924

Gabre -0.49 0.001277 0.027016

St3gal1 0.23 0.001287 0.027187

4631416L12Rik -0.17 0.001292 0.027196

Prom2 -0.54 0.001291 0.027196

Supt4h1 0.14 0.00129 0.027196

Asns 0.11 0.001294 0.02722

Txndc9 0.10 0.001297 0.02725

Rab5b 0.07 0.001308 0.02744

Ngfr -0.47 0.001319 0.027643

Them4 0.18 0.001321 0.027667

Akr1b3 0.12 0.001324 0.027699

Nfib -0.25 0.001326 0.027701

A3galt2 -0.52 0.001343 0.027902

Cmpk2 0.23 0.001343 0.027902

Lgmn 0.17 0.001341 0.027902

Lpin3 -0.49 0.001342 0.027902

Rps6ka5 -0.23 0.001343 0.027902

Gm15608 -0.34 0.001349 0.027948

Myo1e 0.27 0.001347 0.027948

9030617O03Rik -0.18 0.001356 0.028009

Shank1 -0.29 0.001354 0.028009

Uap1 0.18 0.001357 0.028009 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Usp54 -0.18 0.001358 0.028009

Hmgn3 0.11 0.001362 0.028032

Rc3h1 -0.19 0.001361 0.028032

Atp6v0b 0.14 0.001363 0.028033

Hmg20a 0.10 0.001375 0.028204

Lct -0.47 0.001374 0.028204

Gm12977 -0.54 0.001383 0.028303

Myst3 -0.13 0.001383 0.028303

Fubp1 0.15 0.001387 0.028321

Trib2 0.28 0.001385 0.028321

2810004N23Rik 0.15 0.001391 0.028376

4930556I23Rik -0.36 0.001402 0.028516

Cyb561 0.13 0.0014 0.028516

Entpd7 -0.17 0.001402 0.028516

Gm5901 -0.17 0.001408 0.028552

Slc22a4 -0.33 0.001409 0.028552

Zc3hav1 -0.22 0.001409 0.028552

Pde6b -0.51 0.001411 0.028554

Copz1 0.10 0.001415 0.028604

Rnf181 0.10 0.001416 0.028604

Smad9 -0.18 0.001425 0.028751

Eif4e 0.14 0.001431 0.028839

Irgm1 0.22 0.001443 0.029052

Gchfr -0.53 0.001448 0.029124

Git2 -0.18 0.001456 0.029247

Gm11126 -0.32 0.001458 0.029247

Enah -0.15 0.00146 0.029271 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Dusp14 0.22 0.001477 0.029572

7SK.5 0.54 0.001488 0.029755

1810063B07Rik 0.17 0.001494 0.029845

Sec16a -0.09 0.001495 0.029845

Ceacam2 -0.44 0.001503 0.029901

Hprt 0.15 0.001503 0.029901

Qrsl1 0.14 0.001503 0.029901

Hnf1b -0.50 0.001508 0.029903

Mrpl49 0.10 0.001508 0.029903

Snap47 0.10 0.001507 0.029903

3110035E14Rik 0.32 0.001515 0.030007

Arl1 0.12 0.001523 0.030093

Snx15 0.15 0.001522 0.030093

Acsf2 -0.14 0.001529 0.030127

Atg7 0.11 0.001529 0.030127

Tnrc18 -0.21 0.001528 0.030127

9130019P16Rik -0.49 0.001545 0.030394

Atpaf1 0.13 0.00155 0.030406

Ghitm 0.11 0.001549 0.030406

Kif26b -0.30 0.001549 0.030406

Apoe 0.16 0.001554 0.030441

Neb -0.44 0.001561 0.03055

Mboat2 -0.16 0.001567 0.030638

Morf4l2 0.14 0.001571 0.030692

Arpp21 0.15 0.00158 0.030808

Creld1 0.17 0.001581 0.030808

9330154K18Rik -0.52 0.001588 0.030878 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Fzd7 -0.34 0.001591 0.030878

Mrpl19 0.12 0.001589 0.030878

Nr3c2 -0.26 0.001591 0.030878

Sult5a1 -0.53 0.001596 0.030907

Tomm5 0.15 0.001595 0.030907

Gm15631 0.25 0.0016 0.030954

Extl2 0.18 0.001603 0.030971

Gemin8 0.13 0.001608 0.030985

Nexn -0.44 0.001608 0.030985

Tbc1d16 -0.18 0.001607 0.030985

Abcc12 -0.54 0.001615 0.031076

Psenen 0.14 0.001619 0.031114

Mzt1 0.18 0.001624 0.031126

Pdp2 -0.19 0.001622 0.031126

Tmem107 0.24 0.001623 0.031126

Gm16601 -0.26 0.001633 0.031261

Klhl3 -0.22 0.001642 0.03131

Pofut1 -0.14 0.001638 0.03131

Qrfp -0.53 0.001639 0.03131

Trpa1 -0.54 0.001642 0.03131

P2ry12 0.17 0.001645 0.031318

Slc9a3 -0.53 0.00165 0.03138

4930547N16Rik -0.46 0.001654 0.031405

Pdzd11 0.11 0.001653 0.031405

Chst15 -0.10 0.001661 0.031419

Gm1987 -0.54 0.001665 0.031419

Mmachc 0.14 0.001659 0.031419 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Pip4k2b 0.11 0.001661 0.031419

Tbr1 0.25 0.001663 0.031419

Tmem50b 0.11 0.001665 0.031419

Fam83d -0.53 0.001681 0.03169

Hexdc -0.10 0.001691 0.031808

Igf1r -0.18 0.00169 0.031808

Ndufa4 0.13 0.001694 0.031833

2510003E04Rik 0.07 0.001699 0.031882

4930422G04Rik -0.23 0.001712 0.032077

Snrpb2 0.18 0.001712 0.032077

Cycs 0.14 0.001727 0.032324

Comp -0.54 0.001731 0.032359

7SK.312 -0.54 0.001737 0.032443

R3hdm1 0.14 0.001746 0.032565

Ppid 0.13 0.001749 0.032597

Pknox2 0.15 0.001752 0.032623

Olfm2 0.32 0.001756 0.03266

Kctd8 -0.36 0.001776 0.032786

Mcts1 0.16 0.001775 0.032786

Nos1 -0.31 0.001777 0.032786

Pitpnm2 -0.23 0.001774 0.032786

Smarcd2 -0.22 0.001773 0.032786

Sod3 -0.22 0.001768 0.032786

Tardbp 0.15 0.00177 0.032786

Tex11 -0.51 0.001771 0.032786

Mtmr6 0.13 0.001782 0.032809

Stk39 0.13 0.001781 0.032809 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

4930534B04Rik -0.20 0.001788 0.032843

AC118017.1 -0.29 0.001798 0.032843

Barx2 0.32 0.0018 0.032843

Dimt1 0.14 0.001794 0.032843

Elavl4 0.24 0.00179 0.032843

Gpr12 0.23 0.001789 0.032843

Ltbp3 -0.26 0.001795 0.032843

Mapk10 0.13 0.001798 0.032843

Sfrp1 0.33 0.001798 0.032843

Gm15787 -0.30 0.001805 0.032895

AC044864.2 -0.20 0.00181 0.032934

Klf13 -0.21 0.001809 0.032934

Dazap2 0.09 0.001814 0.03297

Wnk4 -0.32 0.001821 0.033057

Col17a1 -0.53 0.001827 0.033134

Megf11 -0.33 0.00183 0.033169

Necap1 0.11 0.001837 0.033254

Alkbh5 -0.15 0.001842 0.033279

Atp6v1d 0.11 0.001841 0.033279

Uba3 0.18 0.001848 0.033347

Cdc123 0.09 0.001855 0.033422

Wnt5b -0.37 0.001855 0.033422

Man2b2 -0.17 0.001861 0.033494

Txndc15 0.09 0.001866 0.033539

Cpt2 -0.18 0.001879 0.033742

SNORA17.514 -0.51 0.001883 0.033784

Atp6ap2 0.16 0.001888 0.033826 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Tfg 0.10 0.001889 0.033826

Gm10925 0.25 0.001891 0.033827

Gm13405 -0.53 0.001898 0.033923

Arntl 0.19 0.001909 0.034

Chn1 0.10 0.001912 0.034

Dctn6 0.10 0.001911 0.034

Hsd17b12 0.11 0.001905 0.034

Mras 0.11 0.001909 0.034

Tmem194 -0.21 0.001917 0.034065

Aanat -0.32 0.001922 0.034125

Pard6g -0.26 0.001933 0.034288

Grin3a -0.15 0.001945 0.034466

Atl1 0.11 0.001949 0.034475

Nxn -0.26 0.001951 0.034475

Uros 0.11 0.001951 0.034475

Rbmxrt 0.11 0.001954 0.034478

A430071A18Rik -0.52 0.001959 0.034547

Golph3 0.12 0.001968 0.034669

Add1 0.08 0.001974 0.034737

Fam107b -0.47 0.001976 0.034737

Tusc2 0.12 0.001992 0.034991

AL929166.1 0.25 0.001996 0.035033

Gm13069 -0.47 0.001999 0.03504

Ppp3r1 0.16 0.002011 0.035221

Kcnk10 -0.33 0.002017 0.035296

9330162012Rik -0.24 0.00202 0.035312

1700015E13Rik -0.45 0.002024 0.035342 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

H2-M10.2 -0.29 0.002042 0.035622

Rps6ka1 -0.35 0.002044 0.035622

Chsy3 0.25 0.002048 0.035626

Gca 0.18 0.002047 0.035626

Fam174a 0.12 0.002051 0.035641

Ndufv2 0.12 0.002062 0.035801

Cerkl -0.37 0.002069 0.035889

1810063B05Rik 0.18 0.002076 0.035911

Gm15722 -0.53 0.002075 0.035911

Tekt4 -0.53 0.002075 0.035911

Gm16952 -0.47 0.002078 0.035917

Ddo -0.29 0.002084 0.035984

L1cam 0.12 0.002093 0.0361

Kit -0.27 0.0021 0.036198

Otx1 0.37 0.002112 0.036346

Sv2c -0.20 0.002113 0.036346

2310040G07Rik -0.43 0.002117 0.036375

Cadps2 -0.38 0.002123 0.036389

Gm11201 -0.47 0.002123 0.036389

Zwint 0.12 0.002122 0.036389

Ppargc1b -0.24 0.002126 0.036394

B3gat1 0.13 0.002143 0.036622

Dpf3 -0.47 0.002142 0.036622

Rere -0.13 0.002147 0.03665

Atp1a4 -0.44 0.002158 0.036817

Itpkb -0.23 0.002164 0.036855

Slc6a5 -0.42 0.002165 0.036855 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Zranb2 0.18 0.002169 0.036894

Kif3b 0.09 0.002179 0.037026

AC163094.1 0.25 0.002208 0.037478

Npb -0.52 0.002209 0.037478

Echdc2 -0.27 0.002221 0.03758

Gatad2b -0.17 0.002225 0.03758

Mrpl40 0.15 0.002226 0.03758

Stk11ip 0.14 0.002218 0.03758

Zfp111 -0.12 0.002226 0.03758

Fam123c -0.15 0.002237 0.037732

Ttc36 -0.50 0.002251 0.037934

Gatc 0.10 0.002261 0.038043

Per3 -0.31 0.00226 0.038043

Bptf -0.11 0.002267 0.038095

Fkbp3 0.16 0.002274 0.038125

Olfr544 -0.52 0.002277 0.038125

Spns3 -0.42 0.002275 0.038125

Vps4a 0.08 0.002272 0.038125

1110002B05Rik 0.14 0.002283 0.038204

Slc22a18 -0.39 0.002287 0.038222

Cpe 0.11 0.002297 0.038357

1110008P14Rik 0.31 0.002303 0.038396

A930001C03Rik -0.52 0.002303 0.038396

Uck1 0.11 0.002311 0.038486

Pxn -0.26 0.002313 0.038486

Gm11738 -0.52 0.002321 0.038582

Klhl18 0.11 0.002331 0.038683 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Lamb2 -0.24 0.002331 0.038683

Clmn -0.17 0.002336 0.03871

Pcdha1 0.12 0.002337 0.03871

Rbm17 0.08 0.00234 0.038732

Fat4 -0.28 0.002344 0.038754

Phyhipl 0.11 0.002347 0.038773

Stard3nl 0.11 0.002354 0.038817

Wls 0.22 0.002353 0.038817

Cebpg 0.14 0.002359 0.038871

Egr1 0.47 0.00238 0.039144

Siah1a 0.11 0.00238 0.039144

Prr12 -0.22 0.002387 0.039227

Slc22a15 -0.22 0.00239 0.039227

Sgtb 0.19 0.002399 0.03935

2310046A06Rik 0.17 0.002417 0.03951

B3gat2 -0.16 0.002415 0.03951

Gm9754 -0.44 0.002419 0.03951

Nfya 0.17 0.002424 0.03951

P2rx1 -0.50 0.002423 0.03951

Pcsk2 0.15 0.00242 0.03951

Wnt6 -0.45 0.002424 0.03951

C1ql2 -0.52 0.002429 0.039521

Pskh1 -0.14 0.002429 0.039521

AI480653 0.15 0.002441 0.039672

Commd3 0.11 0.002459 0.039943

Lyrm1 0.14 0.002463 0.039959

Prx -0.32 0.002487 0.040292 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Tmem127 0.09 0.002486 0.040292

Vps45 0.11 0.00249 0.040294

Srsf6 0.11 0.002495 0.040345

Fkbp10 -0.22 0.002519 0.040656

Sox2 0.21 0.002518 0.040656

Cd163 -0.39 0.00253 0.040751

Gm16267 -0.37 0.002531 0.040751

Tmem170 -0.27 0.002531 0.040751

Cdc42 0.14 0.002544 0.040877

Hdac4 -0.16 0.002542 0.040877

2010002N04Rik -0.33 0.002555 0.040932

Alms1 -0.25 0.002552 0.040932

Tra2a 0.17 0.002556 0.040932

Znrf1 0.14 0.002551 0.040932

Ghrl -0.23 0.002566 0.041023

Sycp2 -0.28 0.002566 0.041023

Rtp1 -0.51 0.002582 0.041232

Ntn1 -0.34 0.002586 0.041267

H2afz 0.11 0.002602 0.041449

Prrt4 -0.31 0.002601 0.041449

Glrx5 0.15 0.002607 0.041501

Arf2 0.10 0.002631 0.041514

BC018473 -0.44 0.002618 0.041514

C130039O16Rik -0.13 0.002614 0.041514

Cldn1 0.38 0.002628 0.041514

Gm15345 -0.50 0.002629 0.041514

Kifc2 0.16 0.002616 0.041514 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Mpp4 -0.41 0.002621 0.041514

Rac1 0.09 0.002619 0.041514

Snx3 0.11 0.002629 0.041514

Wwc2 -0.12 0.002627 0.041514

Fam136a 0.14 0.002635 0.041546

Cdc42bpg -0.37 0.002661 0.041888

Gas2l3 -0.31 0.002664 0.041888

Tspyl1 0.08 0.002662 0.041888

RP23-353F16.2 -0.52 0.002668 0.04192

Psap 0.09 0.002686 0.042132

Serpine1 -0.43 0.002686 0.042132

Mrpl50 0.14 0.002694 0.042223

Zic2 -0.48 0.002704 0.042344

Reep1 0.13 0.002721 0.042562

Pfn2 0.12 0.002731 0.04269

Ebna1bp2 0.09 0.002735 0.042717

Gtf2h5 0.10 0.002742 0.042785

Pfas -0.13 0.002754 0.042863

Rorc -0.36 0.002754 0.042863

Sgsh -0.17 0.002751 0.042863

Adamts14 -0.51 0.002761 0.042897

Polg2 -0.23 0.00276 0.042897

Dhtkd1 -0.15 0.002766 0.042913

Dsg2 0.34 0.002765 0.042913

Plcg2 -0.22 0.00277 0.042939

Zdbf2 -0.30 0.002781 0.043069

Fndc1 -0.30 0.002788 0.043137 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Rap1gap2 0.11 0.002797 0.043241

Hexa 0.13 0.002804 0.043308

Dopey2 -0.10 0.002811 0.043347

Mapk8 0.23 0.00281 0.043347

Rab9 0.10 0.002826 0.043536

1700061G19Rik -0.40 0.002831 0.043579

Syce2 -0.25 0.002835 0.043612

6430562O15Rik -0.32 0.002843 0.043653

Pcna 0.09 0.002843 0.043653

A930028C08Rik -0.32 0.002851 0.043742

4921518J05Rik -0.27 0.002858 0.043816

Atg2b -0.11 0.002863 0.043849

Cox7b 0.13 0.002868 0.043892

1700056N10Rik -0.22 0.002888 0.044118

Acyp1 0.20 0.00289 0.044118

Gm14015 0.51 0.002888 0.044118

Hnrnpm 0.11 0.002893 0.044128

Glt8d2 0.24 0.002897 0.044157

Atp2a3 -0.49 0.002901 0.044176

Bloc1s3 0.16 0.002904 0.044186

Mterfd1 0.15 0.002923 0.044433

Dusp22 0.10 0.002931 0.044527

Gm14169 0.17 0.00295 0.044777

Slc6a7 0.15 0.002959 0.044872

Ccdc78 -0.37 0.002962 0.044889

Akr1b8 -0.37 0.002965 0.04489

Ribc1 -0.37 0.00297 0.044926 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

AC151908.1 -0.17 0.002977 0.044996

Elmod3 -0.13 0.002982 0.045041

Gm10818 -0.35 0.002992 0.045115

Ston1 -0.27 0.00299 0.045115

Mtor -0.10 0.003011 0.045302

Nhsl2 -0.13 0.003012 0.045302

Pyroxd2 -0.25 0.003011 0.045302

1700026J04Rik -0.34 0.003026 0.04548

Uts2 -0.51 0.003038 0.045625

Bbs12 0.23 0.003055 0.045756

Cacng8 -0.28 0.003054 0.045756

P2rx5 -0.44 0.00305 0.045756

Cdca7l 0.28 0.003062 0.045825

Bzw1 0.16 0.003076 0.045998

Btg2 0.37 0.003092 0.046198

1700094C10Rik -0.51 0.003095 0.046203

Plxnb1 -0.16 0.003101 0.046203

Prrg3 0.13 0.003102 0.046203

Vps29 0.13 0.003098 0.046203

Gm16719 -0.51 0.003109 0.046234

Tcp11 -0.40 0.003108 0.046234

Mkl2 -0.14 0.003115 0.046278

Mir665 -0.48 0.003123 0.046362

Dhdpsl -0.35 0.003127 0.046385

Ccdc67 -0.46 0.003137 0.046493

4921536K21Rik -0.43 0.003149 0.046559

Gm5077 -0.40 0.003149 0.046559 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Tns1 -0.17 0.003146 0.046559

Nptn 0.11 0.003174 0.046897

Atp2c2 -0.41 0.003185 0.046927

Pqlc1 -0.18 0.003187 0.046927

Tmem177 0.14 0.00318 0.046927

Zcchc18 0.11 0.003182 0.046927

Tpo -0.50 0.00319 0.04694

Tead4 -0.50 0.003196 0.046989

Hinfp -0.14 0.003212 0.047186

Tcf25 0.09 0.003223 0.04731

Mat2b 0.14 0.003227 0.047317

St3gal6 0.13 0.003229 0.047317

Xrcc3 -0.21 0.003235 0.047381

Snrpd1 0.16 0.003248 0.047529

Cybasc3 0.15 0.003252 0.047541

Syn2 0.09 0.003283 0.047962

9130017K11Rik 0.49 0.003292 0.048037

Elp3 0.08 0.003296 0.048037

Mov10 -0.22 0.003296 0.048037

Abhd2 -0.17 0.003311 0.048222

Wbp4 0.09 0.003318 0.048276

Ccdc25 0.14 0.003323 0.048282

Gpd1l 0.09 0.003325 0.048282

Slc25a4 0.08 0.003326 0.048282

Kdm1a 0.07 0.003331 0.048319

Dffa 0.09 0.00334 0.048376

Ftsjd2 0.10 0.00334 0.048376 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

4930443O20Rik -0.49 0.003366 0.048568

Aurkc -0.49 0.003359 0.048568

Fgfr1op2 0.15 0.003361 0.048568

Mir1942 -0.50 0.003367 0.048568

Trem3 0.50 0.003366 0.048568

Dcxr -0.25 0.003372 0.048608

Bnip3 0.18 0.003382 0.048625

Dusp6 0.36 0.003384 0.048625

Nphs1as -0.50 0.003378 0.048625

Wnt9b -0.49 0.003383 0.048625

Arhgap17 -0.13 0.003389 0.048656

Ank1 -0.31 0.003399 0.048726

Cep164 -0.14 0.003399 0.048726

Gon4l -0.06 0.003416 0.04893

Osbpl1a 0.23 0.003419 0.04894

Abca2 -0.14 0.003423 0.048958

Ppp2r3c 0.16 0.003431 0.049028

Gm15489 -0.32 0.003438 0.049093

Shisa5 0.15 0.003441 0.049093

Abcc5 0.13 0.00345 0.049177

Gm10565 -0.48 0.003452 0.049177

CT030259.3 -0.31 0.003455 0.049179

BB114351 -0.41 0.003465 0.04929

Colq -0.50 0.003478 0.049394

Tnrc6b -0.13 0.003478 0.049394

Actb -0.11 0.003515 0.049877

bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Supplementary Table S2. Differentially expressed genes from the PVC of APOE3/4 vs. APOE3/3 mice

log2 Gene Name p-value FDR FoldChange

Serpina3m 1.86 1.00E-36 1.66E-32

Oscar 1.60 5.49E-29 4.55E-25

Serpina3n 1.53 4.45E-28 2.46E-24

Rdh13 -0.53 1.91E-26 7.90E-23

Ptprh 1.13 1.84E-14 6.09E-11

Thnsl1 0.49 3.19E-12 8.82E-09

Tmc4 0.71 5.55E-11 1.31E-07

Dctd -0.57 9.84E-11 2.04E-07

Wdfy1 -0.54 1.12E-09 2.07E-06

Zc3h7b 0.25 2.11E-09 3.49E-06

Gm15494 -0.81 2.63E-09 3.96E-06

BC060267 -0.46 5.93E-09 8.18E-06

Fmo2 -0.77 1.96E-08 2.49E-05

Gm8759 0.48 2.29E-08 2.71E-05

2210404J11Rik -0.31 3.78E-08 4.17E-05

Serpina3g 0.74 1.63E-07 0.000168

Asprv1 -0.62 4.73E-07 0.00046

RP24-161J8.2 -0.62 7.13E-07 0.000656

Gm1082 0.72 8.29E-07 0.000657

Hif3a -0.67 8.33E-07 0.000657

Psg16 -0.23 7.59E-07 0.000657

Fbxo5 0.59 9.36E-07 0.000705

Ica1 -0.35 1.46E-06 0.001051 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Maml3 0.46 5.88E-06 0.003894

Nrbp2 -0.26 5.76E-06 0.003894

1700048O20Rik -0.46 9.66E-06 0.005928

Gm10931 -0.59 9.33E-06 0.005928

Gm1078 -0.46 1.30E-05 0.007719

Gm15594 -0.64 1.43E-05 0.008191

Pisd -0.28 1.63E-05 0.009001

Gm16854 -0.61 1.77E-05 0.009452

Rps14 0.24 1.94E-05 0.010054

Gm5506 0.39 2.05E-05 0.010293

Gm10094 0.57 2.13E-05 0.010379

Rpa3 -0.34 2.53E-05 0.011949

C8g -0.35 3.31E-05 0.014492

Gm10654 -0.40 3.32E-05 0.014492

Rpl30 0.20 3.16E-05 0.014492

Cdkn2c 0.45 4.27E-05 0.018126

Akt2 0.27 4.82E-05 0.019952

Tcea2 -0.23 4.96E-05 0.020029

Gm9846 0.45 5.45E-05 0.021494

Mtmr4 -0.16 7.52E-05 0.028961

Zfp629 -0.13 8.44E-05 0.031786

Irak1 -0.19 9.10E-05 0.03349

Srp54b 0.27 0.000108 0.039036

Gm10268 0.41 0.000122 0.042921

Nhlh1 -0.56 0.000131 0.044114

Otud1 0.28 0.000129 0.044114

2500003M10Rik 0.13 0.000133 0.044169 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Gm12394 -0.46 0.000139 0.045173

Rps3a 0.19 0.000144 0.045888

Large 0.18 0.000158 0.048456

Nudt3 0.18 0.000158 0.048456

Snrpd3 0.21 0.000161 0.048456

bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Table S3. Differentially expressed untargeted metabolites from the EC of APOE3 vs. E4

mice

Mass (Da) @ Regulation Fold p-value FDR Detection Retention Time (min) in E4/4 Change Mode

[email protected] up 3.25 0.001 0.012 positive [email protected] up 1.48 0.001 0.012 positive [email protected] down 1.31 0.001 0.012 positive [email protected] up 2.72 0.001 0.012 positive [email protected] up 3.55 0.001 0.012 positive [email protected] down 1.38 0.001 0.012 positive [email protected] up 6.27 0.001 0.012 positive [email protected] up 1.66 0.001 0.012 positive [email protected] down 1.55 0.001 0.012 positive [email protected] up 2.26 0.001 0.012 positive [email protected] up 2.78 0.001 0.012 positive [email protected] up 2.25 0.001 0.012 positive [email protected] up 2.43 0.001 0.012 positive [email protected] down 1.33 0.001 0.012 positive [email protected] down 1.28 0.001 0.012 positive [email protected] up 3.75 0.001 0.012 positive [email protected] up 2.05 0.001 0.012 positive [email protected] up 1.47 0.001 0.012 positive [email protected] up 6.65 0.001 0.012 positive [email protected] down 13.97 0.001 0.012 positive [email protected] up 4.07 0.001 0.012 positive [email protected] up 3.37 0.001 0.012 positive [email protected] up 3.85 0.001 0.012 positive [email protected] up 2.49 0.001 0.012 positive [email protected] up 4.04 0.001 0.012 positive [email protected] up 1.98 0.001 0.012 positive [email protected] up 1.87 0.001 0.012 positive [email protected] down 1.78 0.001 0.012 positive [email protected] down 1.24 0.001 0.012 positive [email protected] up 2.73 0.001 0.012 positive [email protected] up 1.63 0.001 0.012 positive [email protected] up 1.74 0.001 0.012 positive [email protected] up 2.25 0.001 0.012 positive [email protected] up 1.84 0.001 0.012 positive [email protected] up 2.89 0.001 0.012 positive bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

[email protected] up 1.68 0.001 0.012 positive [email protected] down 1.41 0.001 0.012 positive [email protected] up 2.48 0.001 0.012 positive [email protected] up 2.25 0.001 0.012 positive [email protected] up 1.66 0.001 0.012 positive [email protected] up 2.40 0.001 0.012 positive [email protected] up 12.77 0.001 0.012 positive [email protected] down 1.54 0.001 0.012 positive [email protected] up 3.65 0.001 0.012 positive [email protected] up 4.48 0.001 0.012 positive [email protected] up 3.62 0.001 0.012 positive [email protected] down 1.35 0.001 0.012 positive [email protected] up 3.49 0.001 0.012 positive [email protected] down 1.29 0.001 0.012 positive [email protected] up 9.38 0.001 0.012 positive [email protected] up 1.35 0.001 0.012 positive [email protected] up 5.83 0.001 0.012 positive [email protected] up 1.87 0.001 0.012 positive [email protected] up 5.47 0.001 0.012 positive [email protected] down 1.97 0.001 0.012 positive [email protected] up 2.59 0.001 0.012 positive [email protected] down 1.70 0.001 0.012 positive [email protected] up 4.02 0.001 0.012 positive [email protected] up 2.08 0.001 0.012 positive [email protected] up 2.22 0.001 0.012 positive [email protected] up 2.56 0.001 0.012 positive [email protected] up 1.58 0.001 0.012 positive [email protected] up 2.85 0.001 0.012 positive [email protected] down 1.43 0.001 0.012 positive [email protected] down 1.43 0.001 0.012 positive [email protected] down 1.45 0.001 0.012 positive [email protected] up 8.56 0.001 0.012 positive [email protected] up 10.13 0.001 0.012 positive [email protected] down 1.60 0.001 0.012 positive [email protected] up 1.79 0.001 0.012 positive [email protected] down 1.22 0.001 0.012 positive [email protected] up 2.43 0.001 0.012 positive [email protected] down 3.14 0.001 0.012 positive [email protected] up 3.73 0.001 0.012 positive [email protected] up 1.72 0.001 0.012 positive [email protected] up 3.77 0.001 0.012 positive bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

[email protected] up 2.33 0.001 0.012 positive [email protected] up 6.45 0.001 0.012 positive [email protected] down 2.16 0.001 0.012 positive [email protected] down 2.18 0.001 0.012 positive [email protected] down 1.64 0.001 0.012 positive [email protected] up 2.05 0.001 0.012 positive [email protected] down 1.74 0.001 0.012 positive [email protected] up 8.86 0.001 0.012 positive [email protected] up 6.10 0.001 0.012 positive [email protected] up 17.67 0.001 0.012 positive [email protected] up 9.18 0.001 0.012 positive [email protected] up 1.63 0.001 0.012 positive [email protected] up 2.84 0.001 0.012 positive [email protected] up 2.72 0.001 0.012 positive [email protected] up 2.26 0.001 0.012 positive [email protected] up 2.66 0.001 0.012 positive [email protected] up 2.26 0.001 0.012 positive [email protected] up 2.67 0.001 0.012 positive [email protected] up 3.74 0.001 0.012 positive [email protected] down 1.78 0.001 0.012 positive [email protected] up 8.95 0.001 0.012 positive [email protected] up 5.47 0.001 0.012 positive [email protected] up 2.96 0.001 0.012 positive [email protected] down 1.90 0.001 0.012 positive [email protected] up 2.23 0.001 0.012 positive [email protected] up 2.71 0.001 0.012 positive [email protected] up 3.02 0.001 0.012 positive [email protected] up 14.72 0.001 0.012 positive [email protected] up 2.14 0.001 0.012 positive [email protected] up 4.17 0.001 0.012 positive [email protected] up 5.36 0.001 0.012 positive [email protected] up 2.58 0.001 0.012 positive [email protected] up 2.08 0.001 0.012 positive [email protected] up 4.08 0.001 0.012 positive [email protected] up 1.86 0.001 0.012 positive [email protected] up 3.31 0.001 0.012 positive [email protected] up 1.80 0.001 0.012 positive [email protected] up 1.24 0.001 0.012 positive [email protected] down 1.70 0.001 0.012 positive [email protected] down 1.56 0.001 0.012 positive [email protected] up 3.10 0.001 0.012 negative bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

[email protected] up 3.19 0.001 0.012 negative [email protected] up 4.25 0.001 0.012 negative [email protected] up 1.62 0.001 0.012 negative [email protected] up 2.17 0.001 0.012 negative [email protected] up 1.66 0.001 0.012 negative [email protected] up 1.86 0.001 0.012 negative [email protected] up 4.37 0.001 0.012 negative [email protected] up 3.08 0.001 0.012 negative [email protected] up 4.10 0.001 0.012 negative [email protected] up 1.81 0.001 0.012 negative [email protected] up 4.48 0.001 0.012 negative [email protected] up 1.59 0.001 0.012 negative [email protected] up 5.40 0.001 0.012 negative [email protected] up 4.27 0.001 0.012 negative [email protected] up 1.66 0.001 0.012 negative [email protected] up 2.60 0.001 0.012 negative [email protected] up 2.55 0.001 0.012 negative [email protected] up 4.87 0.001 0.012 negative [email protected] up 4.10 0.001 0.012 negative [email protected] down 1.44 0.001 0.012 negative [email protected] up 1.56 0.001 0.012 negative [email protected] up 4.61 0.001 0.012 negative [email protected] up 2.62 0.001 0.012 negative [email protected] up 5.60 0.001 0.012 negative [email protected] down 1.42 0.001 0.012 negative [email protected] up 4.78 0.001 0.012 negative [email protected] up 1.50 0.001 0.012 negative [email protected] up 5.98 0.001 0.012 negative [email protected] up 4.71 0.001 0.012 negative [email protected] up 1.81 0.001 0.012 negative [email protected] up 1.61 0.001 0.012 negative [email protected] up 5.60 0.001 0.012 negative [email protected] up 3.36 0.001 0.012 negative [email protected] down 1.25 0.001 0.012 negative [email protected] up 2.16 0.001 0.012 negative [email protected] up 4.98 0.001 0.012 negative [email protected] up 3.20 0.001 0.012 negative [email protected] up 1.74 0.001 0.012 negative [email protected] up 2.97 0.001 0.012 negative [email protected] up 3.57 0.001 0.012 negative [email protected] up 1.36 0.001 0.012 negative bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

[email protected] up 10.65 0.001 0.012 negative [email protected] up 1.62 0.001 0.012 negative [email protected] up 1.94 0.001 0.012 negative [email protected] up 2.04 0.001 0.012 negative [email protected] up 7.45 0.001 0.012 negative [email protected] up 7.30 0.001 0.012 negative [email protected] up 2.04 0.001 0.012 negative [email protected] up 1.46 0.001 0.012 negative [email protected] up 1.96 0.001 0.012 negative [email protected] up 4.26 0.001 0.012 negative [email protected] up 5.00 0.001 0.012 negative [email protected] up 1.61 0.001 0.012 negative [email protected] up 2.04 0.001 0.012 negative [email protected] up 3.62 0.001 0.012 negative [email protected] up 1.89 0.001 0.012 negative [email protected] up 1.68 0.001 0.012 negative [email protected] up 2.22 0.001 0.012 negative [email protected] up 2.22 0.001 0.012 negative [email protected] up 2.03 0.001 0.012 negative [email protected] up 1.49 0.001 0.012 negative [email protected] up 2.29 0.001 0.012 negative [email protected] up 1.50 0.001 0.012 negative [email protected] up 3.56 0.001 0.012 negative [email protected] up 3.35 0.001 0.012 negative [email protected] up 1.53 0.001 0.012 negative [email protected] up 17.11 0.001 0.012 negative [email protected] up 4.37 0.001 0.012 negative [email protected] up 6.40 0.001 0.012 negative [email protected] up 1.83 0.001 0.012 negative [email protected] up 2.00 0.001 0.012 negative [email protected] up 1.67 0.001 0.012 negative [email protected] up 4.94 0.001 0.012 negative [email protected] up 1.72 0.001 0.012 negative [email protected] up 1.51 0.001 0.012 negative [email protected] up 2.57 0.001 0.012 negative [email protected] up 2.56 0.001 0.012 negative [email protected] up 1.97 0.001 0.012 negative [email protected] up 16.89 0.001 0.012 negative [email protected] up 2.53 0.001 0.012 negative [email protected] up 2.44 0.001 0.012 negative [email protected] up 4.24 0.001 0.012 negative bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

[email protected] down 4.38 0.001 0.012 negative [email protected] up 3.81 0.001 0.012 negative [email protected] up 2.37 0.001 0.012 negative [email protected] up 3.07 0.001 0.012 negative [email protected] down 1.26 0.002 0.016 positive [email protected] down 1.33 0.002 0.016 positive [email protected] up 2.11 0.002 0.016 positive [email protected] up 1.73 0.002 0.016 positive [email protected] down 1.55 0.002 0.016 positive [email protected] down 1.56 0.002 0.016 positive [email protected] down 1.72 0.002 0.016 positive [email protected] up 2.87 0.002 0.016 positive [email protected] down 1.37 0.002 0.016 positive [email protected] up 1.65 0.002 0.016 positive [email protected] up 1.27 0.002 0.016 positive [email protected] up 1.97 0.002 0.016 positive [email protected] up 3.06 0.002 0.016 positive [email protected] down 1.36 0.002 0.016 positive [email protected] up 1.50 0.002 0.016 positive [email protected] up 1.71 0.002 0.016 positive [email protected] up 1.96 0.002 0.018 negative [email protected] up 2.12 0.002 0.018 negative [email protected] up 2.37 0.002 0.018 negative [email protected] up 1.74 0.002 0.018 negative [email protected] up 3.98 0.002 0.018 negative [email protected] down 1.33 0.003 0.021 positive [email protected] up 1.38 0.003 0.021 positive [email protected] down 1.34 0.003 0.021 positive [email protected] down 1.23 0.003 0.021 positive [email protected] down 1.23 0.003 0.021 positive [email protected] down 1.38 0.003 0.021 positive [email protected] up 1.71 0.003 0.021 positive [email protected] down 1.53 0.003 0.021 positive [email protected] up 1.98 0.003 0.021 positive [email protected] up 7.14 0.003 0.021 positive [email protected] up 3.05 0.003 0.021 positive [email protected] up 1.93 0.003 0.021 positive [email protected] down 1.31 0.003 0.021 positive [email protected] up 2.69 0.003 0.021 positive [email protected] down 1.28 0.003 0.021 positive [email protected] up 1.31 0.003 0.024 negative bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

[email protected] down 1.27 0.003 0.024 negative [email protected] up 1.82 0.003 0.024 negative [email protected] up 1.31 0.003 0.024 negative [email protected] up 1.22 0.003 0.024 negative [email protected] up 2.89 0.003 0.024 negative [email protected] up 4.19 0.003 0.024 negative [email protected] up 2.02 0.003 0.024 negative [email protected] down 1.65 0.003 0.024 negative [email protected] up 1.25 0.004 0.028 positive [email protected] down 17.74 0.004 0.028 positive [email protected] up 14.61 0.004 0.028 positive [email protected] down 1.36 0.004 0.028 positive [email protected] up 1.66 0.004 0.028 positive [email protected] down 1.69 0.004 0.028 positive [email protected] up 1.61 0.004 0.028 positive [email protected] up 1.86 0.004 0.028 positive [email protected] up 1.48 0.004 0.028 positive [email protected] down 1.27 0.004 0.028 positive [email protected] up 1.66 0.004 0.028 positive [email protected] down 1.17 0.004 0.028 positive [email protected] down 1.42 0.004 0.032 negative [email protected] down 1.37 0.004 0.032 negative [email protected] down 1.63 0.004 0.032 negative [email protected] up 1.33 0.004 0.032 negative [email protected] up 1.95 0.004 0.032 negative [email protected] up 1.81 0.004 0.032 negative [email protected] up 2.04 0.004 0.032 negative [email protected] down 4.37 0.005 0.043 negative [email protected] up 1.82 0.005 0.036 positive [email protected] up 1.62 0.005 0.036 positive [email protected] up 1.75 0.005 0.036 positive [email protected] up 1.22 0.005 0.036 positive [email protected] up 2.56 0.005 0.036 positive [email protected] up 1.80 0.005 0.036 positive [email protected] up 1.72 0.005 0.036 positive [email protected] up 1.49 0.005 0.036 positive [email protected] down 1.36 0.005 0.036 positive [email protected] up 1.76 0.005 0.036 positive [email protected] down 1.41 0.005 0.036 positive [email protected] up 2.42 0.005 0.036 positive [email protected] down 2.74 0.005 0.036 positive bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

[email protected] down 1.51 0.005 0.036 positive [email protected] down 1.34 0.005 0.036 positive [email protected] up 5.91 0.005 0.036 positive [email protected] up 1.15 0.005 0.043 negative [email protected] up 1.37 0.005 0.043 negative [email protected] up 3.96 0.005 0.043 negative [email protected] down 1.41 0.005 0.043 negative [email protected] up 1.15 0.005 0.043 negative [email protected] up 2.02 0.005 0.043 negative [email protected] up 1.34 0.005 0.043 negative [email protected] up 2.33 0.005 0.043 negative [email protected] up 1.87 0.008 0.048 positive [email protected] down 1.21 0.008 0.048 positive [email protected] down 1.18 0.008 0.048 positive [email protected] up 1.98 0.008 0.048 positive [email protected] down 1.19 0.008 0.048 positive [email protected] up 1.94 0.008 0.048 positive [email protected] up 1.59 0.008 0.048 positive [email protected] up 2.87 0.008 0.048 positive [email protected] up 2.05 0.008 0.048 positive [email protected] down 1.21 0.008 0.048 positive [email protected] down 1.79 0.008 0.048 positive [email protected] up 2.77 0.008 0.048 positive

bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Table S4. Differentially expressed untargeted metabolites from the PVC of APOE3 vs. E4

mice

Mass (Da) @ Regulation Fold p-value FDR Detection Retention Time (min) in E4/4 Change Mode

[email protected] up 3.11 0.001 0.014 positive [email protected] up 1.51 0.001 0.014 positive [email protected] up 2.64 0.001 0.014 positive [email protected] up 3.55 0.001 0.014 positive [email protected] down 1.37 0.001 0.014 positive [email protected] up 6.08 0.001 0.014 positive [email protected] up 1.64 0.001 0.014 positive [email protected] down 1.43 0.001 0.014 positive [email protected] up 2.29 0.001 0.014 positive [email protected] up 2.60 0.001 0.014 positive [email protected] up 2.10 0.001 0.014 positive [email protected] up 2.15 0.001 0.014 positive [email protected] down 1.30 0.001 0.014 positive [email protected] up 3.69 0.001 0.014 positive [email protected] up 2.02 0.001 0.014 positive [email protected] up 1.51 0.001 0.014 positive [email protected] up 20.46 0.001 0.014 positive [email protected] up 5.15 0.001 0.014 positive [email protected] down 12.38 0.001 0.014 positive [email protected] down 1.27 0.001 0.014 positive [email protected] up 2.90 0.001 0.014 positive [email protected] up 3.89 0.001 0.014 positive [email protected] up 4.48 0.001 0.014 positive [email protected] up 2.81 0.001 0.014 positive [email protected] up 2.13 0.001 0.014 positive [email protected] up 3.64 0.001 0.014 positive [email protected] up 1.94 0.001 0.014 positive [email protected] down 1.59 0.001 0.014 positive [email protected] down 1.28 0.001 0.014 positive [email protected] up 8.45 0.001 0.014 positive [email protected] up 1.60 0.001 0.014 positive [email protected] up 2.17 0.001 0.014 positive [email protected] up 1.64 0.001 0.014 positive [email protected] up 2.57 0.001 0.014 positive [email protected] up 1.74 0.001 0.014 positive bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

[email protected] down 1.36 0.001 0.014 positive [email protected] up 2.50 0.001 0.014 positive [email protected] up 1.14 0.001 0.014 positive [email protected] up 2.10 0.001 0.014 positive [email protected] up 1.68 0.001 0.014 positive [email protected] up 5.28 0.001 0.014 positive [email protected] up 10.49 0.001 0.014 positive [email protected] down 1.51 0.001 0.014 positive [email protected] up 2.75 0.001 0.014 positive [email protected] down 1.29 0.001 0.014 positive [email protected] up 3.62 0.001 0.014 positive [email protected] up 5.61 0.001 0.014 positive [email protected] up 1.76 0.001 0.014 positive [email protected] up 4.25 0.001 0.014 positive [email protected] down 1.96 0.001 0.014 positive [email protected] up 1.72 0.001 0.014 positive [email protected] down 2.26 0.001 0.014 positive [email protected] up 4.51 0.001 0.014 positive [email protected] up 1.81 0.001 0.014 positive [email protected] up 2.14 0.001 0.014 positive [email protected] up 1.70 0.001 0.014 positive [email protected] up 2.29 0.001 0.014 positive [email protected] up 1.61 0.001 0.014 positive [email protected] up 2.74 0.001 0.014 positive [email protected] down 1.53 0.001 0.014 positive [email protected] down 1.40 0.001 0.014 positive [email protected] up 6.49 0.001 0.014 positive [email protected] up 10.18 0.001 0.014 positive [email protected] down 1.54 0.001 0.014 positive [email protected] up 2.70 0.001 0.014 positive [email protected] up 2.65 0.001 0.014 positive [email protected] up 2.09 0.001 0.014 positive [email protected] down 1.27 0.001 0.014 positive [email protected] up 7.49 0.001 0.014 positive [email protected] down 2.10 0.001 0.014 positive [email protected] down 2.07 0.001 0.014 positive [email protected] up 1.80 0.001 0.014 positive [email protected] up 8.14 0.001 0.014 positive [email protected] up 14.76 0.001 0.014 positive [email protected] up 3.58 0.001 0.014 positive [email protected] up 1.68 0.001 0.014 positive bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

[email protected] up 3.15 0.001 0.014 positive [email protected] up 1.72 0.001 0.014 positive [email protected] up 2.12 0.001 0.014 positive [email protected] up 2.37 0.001 0.014 positive [email protected] up 2.96 0.001 0.014 positive [email protected] up 1.38 0.001 0.014 positive [email protected] down 1.69 0.001 0.014 positive [email protected] up 11.80 0.001 0.014 positive [email protected] up 3.63 0.001 0.014 positive [email protected] up 4.24 0.001 0.014 positive [email protected] up 3.52 0.001 0.014 positive [email protected] up 5.64 0.001 0.014 positive [email protected] up 3.15 0.001 0.014 positive [email protected] up 1.42 0.001 0.014 positive [email protected] up 13.54 0.001 0.014 positive [email protected] up 2.44 0.001 0.014 positive [email protected] up 3.55 0.001 0.014 positive [email protected] up 4.23 0.001 0.014 positive [email protected] up 3.77 0.001 0.014 positive [email protected] up 1.88 0.001 0.014 positive [email protected] up 2.29 0.001 0.014 positive [email protected] up 4.64 0.001 0.014 positive [email protected] up 1.40 0.002 0.018 positive [email protected] up 2.34 0.002 0.018 positive [email protected] up 1.36 0.002 0.018 positive [email protected] up 1.49 0.002 0.018 positive [email protected] down 9.41 0.002 0.018 positive [email protected] down 1.31 0.002 0.018 positive [email protected] down 1.97 0.002 0.018 positive [email protected] up 1.48 0.002 0.018 positive [email protected] down 1.33 0.002 0.018 positive [email protected] down 1.39 0.002 0.018 positive [email protected] up 2.81 0.002 0.018 positive [email protected] down 1.44 0.002 0.018 positive [email protected] up 2.10 0.002 0.018 positive [email protected] up 1.94 0.002 0.018 positive [email protected] down 1.26 0.002 0.018 positive [email protected] up 2.80 0.002 0.018 positive [email protected] up 1.53 0.002 0.018 positive [email protected] down 1.22 0.003 0.023 positive [email protected] down 1.15 0.003 0.023 positive bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

[email protected] up 2.27 0.003 0.023 positive [email protected] down 1.18 0.003 0.023 positive [email protected] up 1.16 0.003 0.023 positive [email protected] down 1.26 0.003 0.023 positive [email protected] up 1.22 0.003 0.023 positive [email protected] down 1.27 0.003 0.023 positive [email protected] up 3.05 0.003 0.023 positive [email protected] up 2.03 0.003 0.023 positive [email protected] up 1.39 0.003 0.023 positive [email protected] down 1.55 0.003 0.023 positive [email protected] up 1.87 0.003 0.023 positive [email protected] up 2.73 0.003 0.023 positive [email protected] up 2.01 0.003 0.023 positive [email protected] up 3.49 0.003 0.023 positive [email protected] up 3.31 0.001 0.026 negative [email protected] up 3.50 0.001 0.026 negative [email protected] up 4.12 0.001 0.026 negative [email protected] up 2.73 0.001 0.026 negative [email protected] up 4.79 0.001 0.026 negative [email protected] up 3.06 0.001 0.026 negative [email protected] up 4.46 0.001 0.026 negative [email protected] up 7.22 0.001 0.026 negative [email protected] up 5.75 0.001 0.026 negative [email protected] up 4.91 0.001 0.026 negative [email protected] up 3.42 0.001 0.026 negative [email protected] up 4.02 0.001 0.026 negative [email protected] up 3.24 0.001 0.026 negative [email protected] up 3.76 0.001 0.026 negative [email protected] up 4.59 0.001 0.026 negative [email protected] up 6.11 0.001 0.026 negative [email protected] up 3.60 0.001 0.026 negative [email protected] up 7.68 0.001 0.026 negative [email protected] up 4.04 0.001 0.026 negative [email protected] up 3.61 0.001 0.026 negative [email protected] up 1.90 0.001 0.026 negative [email protected] up 4.53 0.001 0.026 negative [email protected] up 4.89 0.001 0.026 negative [email protected] up 11.42 0.001 0.026 negative [email protected] up 4.45 0.001 0.026 negative [email protected] up 2.28 0.001 0.026 negative [email protected] up 3.39 0.001 0.026 negative bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

[email protected] up 4.25 0.001 0.026 negative [email protected] up 14.40 0.001 0.026 negative [email protected] up 3.49 0.001 0.026 negative [email protected] up 6.00 0.001 0.026 negative [email protected] up 4.11 0.001 0.026 negative [email protected] down 3.46 0.001 0.026 negative [email protected] up 5.05 0.001 0.026 negative [email protected] down 2.58 0.001 0.026 negative [email protected] up 3.19 0.001 0.026 negative [email protected] up 4.02 0.001 0.026 negative [email protected] up 3.58 0.001 0.026 negative [email protected] up 4.88 0.001 0.026 negative [email protected] up 2.26 0.001 0.026 negative [email protected] up 3.44 0.001 0.026 negative [email protected] up 1.72 0.002 0.027 negative [email protected] down 1.53 0.002 0.027 negative [email protected] up 2.03 0.002 0.027 negative [email protected] down 1.56 0.002 0.027 negative [email protected] up 1.93 0.002 0.027 negative [email protected] up 2.95 0.002 0.027 negative [email protected] down 5.15 0.002 0.027 negative [email protected] up 2.28 0.002 0.027 negative [email protected] up 7.49 0.002 0.027 negative [email protected] up 1.31 0.002 0.027 negative [email protected] up 3.11 0.002 0.027 negative [email protected] up 1.93 0.002 0.027 negative [email protected] up 2.29 0.002 0.027 negative [email protected] up 1.78 0.002 0.027 negative [email protected] up 2.50 0.002 0.027 negative [email protected] up 2.70 0.002 0.027 negative [email protected] up 1.96 0.002 0.027 negative [email protected] up 1.92 0.002 0.027 negative [email protected] up 2.75 0.002 0.027 negative [email protected] up 1.95 0.002 0.027 negative [email protected] up 1.99 0.003 0.030 negative [email protected] up 1.95 0.003 0.030 negative [email protected] up 1.88 0.003 0.030 negative [email protected] up 2.43 0.003 0.030 negative [email protected] up 1.96 0.003 0.030 negative [email protected] up 1.05 0.003 0.030 negative [email protected] up 3.07 0.003 0.030 negative bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

[email protected] up 2.45 0.003 0.030 negative [email protected] up 1.99 0.003 0.030 negative [email protected] down 2.55 0.003 0.030 negative [email protected] up 19.03 0.003 0.030 negative [email protected] up 1.79 0.003 0.030 negative [email protected] up 1.69 0.003 0.030 negative [email protected] up 2.12 0.003 0.030 negative [email protected] up 1.77 0.003 0.030 negative [email protected] up 1.77 0.003 0.030 negative [email protected] down 1.39 0.003 0.030 negative [email protected] up 2.70 0.003 0.030 negative [email protected] down 1.10 0.004 0.031 positive [email protected] down 1.72 0.004 0.031 positive [email protected] down 1.78 0.004 0.031 positive [email protected] down 1.18 0.004 0.031 positive [email protected] up 1.73 0.004 0.031 positive [email protected] up 2.39 0.004 0.031 positive [email protected] down 1.45 0.004 0.031 positive [email protected] down 1.42 0.004 0.031 positive [email protected] up 1.50 0.004 0.031 positive [email protected] up 1.31 0.004 0.031 positive [email protected] down 1.39 0.004 0.031 positive [email protected] down 1.59 0.004 0.031 positive [email protected] down 1.16 0.004 0.031 positive [email protected] up 1.95 0.004 0.039 negative [email protected] up 2.38 0.004 0.039 negative [email protected] up 1.05 0.004 0.039 negative [email protected] down 1.21 0.004 0.039 negative [email protected] up 1.87 0.004 0.039 negative [email protected] up 1.81 0.004 0.039 negative [email protected] down 1.07 0.004 0.039 negative [email protected] up 1.64 0.004 0.039 negative [email protected] up 1.95 0.004 0.039 negative [email protected] up 1.68 0.004 0.039 negative [email protected] down 1.17 0.004 0.039 negative [email protected] down 1.36 0.005 0.041 positive [email protected] down 1.77 0.005 0.041 positive [email protected] up 1.80 0.005 0.041 positive [email protected] up 1.07 0.005 0.041 positive [email protected] up 1.18 0.005 0.041 positive [email protected] up 1.26 0.005 0.041 positive bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

[email protected] down 1.29 0.005 0.041 positive [email protected] down 2.69 0.005 0.041 positive [email protected] down 1.33 0.005 0.041 positive [email protected] up 1.67 0.005 0.041 positive [email protected] up 39.98 0.005 0.046 negative [email protected] up 62.14 0.005 0.046 negative [email protected] up 2.57 0.005 0.046 negative [email protected] up 1.80 0.005 0.046 negative [email protected] up 2.09 0.005 0.046 negative [email protected] up 21.38 0.005 0.046 negative [email protected] up 1.90 0.005 0.046 negative [email protected] up 6.12 0.005 0.046 negative [email protected] up 16.93 0.005 0.046 negative [email protected] down 1.59 0.005 0.046 negative [email protected] up 12.87 0.005 0.046 negative [email protected] up 2.15 0.005 0.046 negative [email protected] up 6.03 0.005 0.046 negative [email protected] down 1.50 0.005 0.046 negative [email protected] up 4.61 0.005 0.046 negative [email protected] up 1.91 0.005 0.046 negative [email protected] up 2.47 0.005 0.046 negative [email protected] up 3.10 0.005 0.046 negative [email protected] up 4.49 0.005 0.046 negative

bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

Supplementary Table S5. Putative identities of untargeted metabolites from the EC of

APOE3 vs. E4 mice by PIUMet analysis

monisotopic- PIUMet- Chemical- mz-peak Metabolite-Name HMDB-ID molecular ID formula weight m/z=82.0441 2-Methylfuran Met20304 HMDB13749 82.04186481 C5H6O m/z=82.0441 2,4-Pentadienal Met11526 HMDB31597 82.04186481 C5H6O m/z=828.2752 Maltopentaose Met40535 HMDB12254 828.2746818 C30H52O26 m/z=88.0534 Butyric acid Met4442 HMDB00039 88.0524295 C4H8O2 m/z=88.0534 Butyric acid Met4435 HMDB00039 88.0524295 C4H8O2 m/z=88.0534 Butyric acid Met4437 HMDB00039 88.0524295 C4H8O2 m/z=88.0534 Isobutyric acid Met34213 HMDB01873 88.0524295 C4H8O2 m/z=88.0534 Methyl propionate Met26873 HMDB30062 88.0524295 C4H8O2 m/z=88.0534 Ethyl acetate Met42630 HMDB31217 88.0524295 C4H8O2 m/z=88.0534 1-Hydroxy-2-butanone Met27352 HMDB31507 88.0524295 C4H8O2 m/z=88.0534 Propyl formate Met12858 HMDB40253 88.0524295 C4H8O2 m/z=88.0534 Isopropyl formate Met11917 HMDB40579 88.0524295 C4H8O2 m/z=90.0304 L-Lactic acid Met2685 HMDB00190 90.03169406 C3H6O3 m/z=90.0304 L-Lactic acid Met2690 HMDB00190 90.03169406 C3H6O3 m/z=90.0304 L-Lactic acid Met2687 HMDB00190 90.03169406 C3H6O3 m/z=90.0304 L-Lactic acid Met2688 HMDB00190 90.03169406 C3H6O3 m/z=90.0304 Hydroxypropionic acid Met395 HMDB00700 90.03169406 C3H6O3 m/z=90.0304 Hydroxypropionic acid Met383 HMDB00700 90.03169406 C3H6O3 m/z=90.0304 Glyceraldehyde Met2434 HMDB01051 90.03169406 C3H6O3 m/z=90.0304 Glyceraldehyde Met2900 HMDB01051 90.03169406 C3H6O3 m/z=90.0304 D-Lactic acid Met77 HMDB01311 90.03169406 C3H6O3 m/z=90.0304 D-Lactic acid Met71 HMDB01311 90.03169406 C3H6O3 m/z=90.0304 D-Lactic acid Met69 HMDB01311 90.03169406 C3H6O3 m/z=90.0304 Dihydroxyacetone Met3499 HMDB01882 90.03169406 C3H6O3 m/z=90.0304 Dihydroxyacetone Met386 HMDB01882 90.03169406 C3H6O3 m/z=90.0304 Dimethyl carbonate Met15485 HMDB29580 90.03169406 C3H6O3 m/z=90.0304 Monoethyl carbonate Met13334 HMDB31232 90.03169406 C3H6O3 m/z=90.0304 Methoxyacetic acid Met38392 HMDB41929 90.03169406 C3H6O3 m/z=104.0436 2-Hydroxybutyric acid Met1435 HMDB00008 104.0473441 C4H8O3 m/z=104.0436 2-Hydroxybutyric acid Met1439 HMDB00008 104.0473441 C4H8O3 m/z=104.0436 (R)-3-Hydroxybutyric acid Met837 HMDB00011 104.0473441 C4H8O3 m/z=104.0436 (R)-3-Hydroxybutyric acid Met831 HMDB00011 104.0473441 C4H8O3 m/z=104.0436 (R)-3-Hydroxybutyric acid Met833 HMDB00011 104.0473441 C4H8O3 m/z=104.0436 (S)-3-Hydroxyisobutyric acid Met1299 HMDB00023 104.0473441 C4H8O3 m/z=104.0436 (R)-3-Hydroxyisobutyric acid Met41554 HMDB00336 104.0473441 C4H8O3 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

m/z=104.0436 3-Hydroxybutyric acid Met31206 HMDB00357 104.0473441 C4H8O3 m/z=104.0436 (S)-3-Hydroxybutyric acid Met34843 HMDB00442 104.0473441 C4H8O3 m/z=104.0436 4-Hydroxybutyric acid Met38085 HMDB00710 104.0473441 C4H8O3 m/z=104.0436 Ethoxyacetic acid Met9767 HMDB31212 104.0473441 C4H8O3 m/z=116.0607 2-Methyl-1-methylthio-2-butene Met34744 HMDB32411 116.0659711 C6H12S m/z=116.0607 1-Propenyl propyl sulfide Met37762 HMDB40244 116.0659711 C6H12S m/z=118.0257 Methylmalonic acid Met39066 HMDB00202 118.0266087 C4H6O4 m/z=118.0257 Succinic acid Met3924 HMDB00254 118.0266087 C4H6O4 m/z=118.0257 Succinic acid Met2481 HMDB00254 118.0266087 C4H6O4 m/z=118.0257 Succinic acid Met433 HMDB00254 118.0266087 C4H6O4 m/z=118.0257 Succinic acid Met434 HMDB00254 118.0266087 C4H6O4 m/z=118.0257 Succinic acid Met436 HMDB00254 118.0266087 C4H6O4 m/z=118.0257 Erythrono-1,4-lactone Met6527 HMDB00349 118.0266087 C4H6O4 m/z=118.0257 Threonolactone Met18515 HMDB00940 118.0266087 C4H6O4 m/z=118.0257 4-Hydroxy-2-oxobutanoic acid Met27915 HMDB31204 118.0266087 C4H6O4 m/z=118.0257 xi-3-Hydroxy-2-oxobutanoic acid Met19871 HMDB39324 118.0266087 C4H6O4 m/z=118.0634 3-Hydroxy-2-methyl-[R-(R,S)]-butanoic acid Met31201 HMDB00351 118.0629942 C5H10O3 m/z=118.0634 2-Methyl-3-hydroxybutyric acid Met31203 HMDB00354 118.0629942 C5H10O3 m/z=118.0634 2-Ethylhydracrylic acid Met13611 HMDB00396 118.0629942 C5H10O3 m/z=118.0634 2-Hydroxy-3-methylbutyric acid Met23796 HMDB00407 118.0629942 C5H10O3 m/z=118.0634 3-Hydroxy-2-methyl-[S-(R,R)]-butanoic acid Met8473 HMDB00410 118.0629942 C5H10O3 m/z=118.0634 3-Hydroxyvaleric acid Met18759 HMDB00531 118.0629942 C5H10O3 m/z=118.0634 Erythronilic acid Met44489 HMDB00642 118.0629942 C5H10O3 m/z=118.0634 3-Hydroxyisovaleric acid Met19003 HMDB00754 118.0629942 C5H10O3 m/z=118.0634 2-Hydroxyvaleric acid Met9566 HMDB01863 118.0629942 C5H10O3 m/z=118.0634 2-Hydroxy-2-methylbutyric acid Met12795 HMDB01987 118.0629942 C5H10O3 m/z=118.0634 4-Hydroxyisovaleric acid Met18300 HMDB02011 118.0629942 C5H10O3 m/z=118.0634 Ethyl lactate Met26153 HMDB40735 118.0629942 C5H10O3 m/z=118.0634 Methyl 3-hydroxybutyrate Met40266 HMDB41603 118.0629942 C5H10O3 m/z=118.0635 3-Hydroxy-2-methyl-[R-(R,S)]-butanoic acid Met31201 HMDB00351 118.0629942 C5H10O3 m/z=118.0635 2-Methyl-3-hydroxybutyric acid Met31203 HMDB00354 118.0629942 C5H10O3 m/z=118.0635 2-Ethylhydracrylic acid Met13611 HMDB00396 118.0629942 C5H10O3 m/z=118.0635 2-Hydroxy-3-methylbutyric acid Met23796 HMDB00407 118.0629942 C5H10O3 m/z=118.0635 3-Hydroxy-2-methyl-[S-(R,R)]-butanoic acid Met8473 HMDB00410 118.0629942 C5H10O3 m/z=118.0635 3-Hydroxyvaleric acid Met18759 HMDB00531 118.0629942 C5H10O3 m/z=118.0635 Erythronilic acid Met44489 HMDB00642 118.0629942 C5H10O3 m/z=118.0635 3-Hydroxyisovaleric acid Met19003 HMDB00754 118.0629942 C5H10O3 m/z=118.0635 2-Hydroxyvaleric acid Met9566 HMDB01863 118.0629942 C5H10O3 m/z=118.0635 2-Hydroxy-2-methylbutyric acid Met12795 HMDB01987 118.0629942 C5H10O3 m/z=118.0635 4-Hydroxyisovaleric acid Met18300 HMDB02011 118.0629942 C5H10O3 m/z=118.0635 Ethyl lactate Met26153 HMDB40735 118.0629942 C5H10O3 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

m/z=118.0635 Methyl 3-hydroxybutyrate Met40266 HMDB41603 118.0629942 C5H10O3 m/z=120.0888 1,2,4-Trimethylbenzene Met23024 HMDB13733 120.0939004 C9H12 m/z=120.0888 Isopropylbenzene Met33772 HMDB34029 120.0939004 C9H12 m/z=120.0888 1,2,4-Tris(methylene)cyclohexane Met43086 HMDB40458 120.0939004 C9H12 m/z=136.0728 Tetrahydropteridine Met19291 HMDB01216 136.0748963 C6H8N4 m/z=136.0728 N-Methylnicotinamide Met42986 HMDB03152 136.0636629 C7H8N2O m/z=136.0728 2-Methylerythritol Met28115 HMDB11659 136.0735589 C5H12O4 m/z=136.0728 2-Acetyl-3-methylpyrazine Met18135 HMDB30001 136.0636629 C7H8N2O m/z=136.0728 3-Deoxy-D-arabinitol Met27920 HMDB31201 136.0735589 C5H12O4 m/z=136.0728 2-Deoxy-D-ribitol Met12863 HMDB33919 136.0735589 C5H12O4 m/z=136.0728 2-Aminobenzamide Met8451 HMDB33947 136.0636629 C7H8N2O m/z=136.0728 4-Acetyl-2-methylpyrimidine Met28071 HMDB37819 136.0636629 C7H8N2O m/z=136.0728 1-Deoxy-D-ribitol Met18415 HMDB41486 136.0735589 C5H12O4 m/z=150.0205 Tartaric acid Met43356 HMDB00956 150.0164379 C4H6O6 m/z=150.0205 2-(2-Thienyl)furan Met13520 HMDB29715 150.0139355 C8H6OS m/z=150.0205 D-Tartaric acid Met18901 HMDB29878 150.0164379 C4H6O6 m/z=150.0205 2-Methyl-2-(methyldithio)propanal Met43729 HMDB38889 150.0173063 C5H10OS2 m/z=150.0877 6-Methylnicotinamide Met14204 HMDB13704 150.079313 C8H10N2O m/z=150.0877 2-Acetyl-3-ethylpyrazine Met26675 HMDB29410 150.079313 C8H10N2O 7,8-Dihydro-3-methylpyrrolo[1,2- m/z=150.0877 Met23285 HMDB38513 150.079313 C8H10N2O a]pyrimidin-2(6H)-one m/z=150.0877 2-Acetyl-3,6-dimethylpyrazine Met27431 HMDB39998 150.079313 C8H10N2O m/z=150.0877 2-Acetyl-3,5-dimethylpyrazine Met27432 HMDB39999 150.079313 C8H10N2O m/z=150.1352 Cystophorene Met15805 HMDB30944 150.1408506 C11H18 m/z=150.1352 (E)-4,8-Dimethyl-1,3,7-nonatriene Met13279 HMDB35792 150.1408506 C11H18 m/z=152.1151 (-)-trans-Carveol Met2940 HMDB03450 152.1201151 C10H16O m/z=152.1151 (-)-trans-Carveol Met2934 HMDB03450 152.1201151 C10H16O m/z=152.1151 Alpha-Pinene-oxide Met3747 HMDB03667 152.1201151 C10H16O m/z=152.1151 Alpha-Pinene-oxide Met3745 HMDB03667 152.1201151 C10H16O m/z=152.1151 p-Menth-3-en-9-al Met9390 HMDB30892 152.1201151 C10H16O m/z=152.1151 Isocitral Met9391 HMDB30893 152.1201151 C10H16O m/z=152.1151 3-Methyl-5-propyl-2-cyclohexen-1-one Met23042 HMDB31854 152.1201151 C10H16O m/z=152.1151 2,6-Dimethyl-1,7-octadien-3-one Met21047 HMDB32120 152.1201151 C10H16O m/z=152.1151 (E)-3,7-Dimethyl-1,5,7-octatrien-3-ol Met41157 HMDB32242 152.1201151 C10H16O m/z=152.1151 2-(trans-2-Pentenyl)cyclopentanone Met26593 HMDB32536 152.1201151 C10H16O m/z=152.1151 p-Mentha-1(6),8-dien-3-ol Met24158 HMDB34715 152.1201151 C10H16O m/z=152.1151 cis-2-Thujen-4-ol Met39498 HMDB34916 152.1201151 C10H16O m/z=152.1151 (R)-Campholenic aldehyde Met29100 HMDB34973 152.1201151 C10H16O m/z=152.1151 (R)-Carvotanacetone Met29099 HMDB34974 152.1201151 C10H16O m/z=152.1151 Piperitone Met29098 HMDB34975 152.1201151 C10H16O m/z=152.1151 (+)-Fenchone Met26109 HMDB34985 152.1201151 C10H16O m/z=152.1151 (1S,4R)-p-Mentha-2,8-dien-1-ol Met23627 HMDB35054 152.1201151 C10H16O bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

m/z=152.1151 (1R,4R)-p-Mentha-2,8-dien-1-ol Met16263 HMDB35077 152.1201151 C10H16O m/z=152.1151 Geranial Met13426 HMDB35078 152.1201151 C10H16O m/z=152.1151 cis-Citral Met44109 HMDB35092 152.1201151 C10H16O m/z=152.1151 2-Pinen-10-ol Met15975 HMDB35100 152.1201151 C10H16O m/z=152.1151 Dehydrolinalool Met26053 HMDB35120 152.1201151 C10H16O m/z=152.1151 Darwinol Met26048 HMDB35125 152.1201151 C10H16O m/z=152.1151 1,2-Epoxy-p-menth-8-ene Met21693 HMDB35158 152.1201151 C10H16O m/z=152.1151 (S)-9,10-Cyclo-p-menth-1-en-4-ol Met12817 HMDB35203 152.1201151 C10H16O (Z)-2-Methyl-6-methylene-2,7-octadien-1- m/z=152.1151 Met33119 HMDB35241 152.1201151 C10H16O ol m/z=152.1151 (2S,4R)-p-Mentha-1(7),5-dien-2-ol Met18660 HMDB35277 152.1201151 C10H16O m/z=152.1151 Teresantalol Met11534 HMDB35280 152.1201151 C10H16O m/z=152.1151 (S)-Phellandral Met38766 HMDB35600 152.1201151 C10H16O m/z=152.1151 (R)-p-Menth-4(8)-en-3-one Met38762 HMDB35604 152.1201151 C10H16O m/z=152.1151 (+)-cis-Carveol Met9427 HMDB35622 152.1201151 C10H16O m/z=152.1151 3-Pinanone Met32762 HMDB35656 152.1201151 C10H16O m/z=152.1151 alpha-Cyclocitral Met26503 HMDB35706 152.1201151 C10H16O m/z=152.1151 (R)-Piperitone Met41265 HMDB35737 152.1201151 C10H16O m/z=152.1151 (S)-Piperitone Met41258 HMDB35738 152.1201151 C10H16O m/z=152.1151 (-)-trans-Isopulegone Met7493 HMDB35741 152.1201151 C10H16O m/z=152.1151 Pinocarveol Met7491 HMDB35743 152.1201151 C10H16O m/z=152.1151 (-)-Isopinocamphone Met7498 HMDB35744 152.1201151 C10H16O m/z=152.1151 Marmelo oxide A Met33868 HMDB36044 152.1201151 C10H16O m/z=152.1151 Dihydrocarvone Met25035 HMDB36079 152.1201151 C10H16O m/z=152.1151 (1S,4S)-Dihydrocarvone Met24059 HMDB36080 152.1201151 C10H16O m/z=152.1151 (-)-cis-Carveol Met7749 HMDB36082 152.1201151 C10H16O m/z=152.1151 Carveol Met7748 HMDB36083 152.1201151 C10H16O m/z=152.1151 Dehydro-1,8-cineole Met7750 HMDB36085 152.1201151 C10H16O m/z=152.1151 (R)-p-Mentha-1,8-dien-7-ol Met7752 HMDB36087 152.1201151 C10H16O m/z=152.1151 (S)-p-Mentha-1,8-dien-7-ol Met7744 HMDB36088 152.1201151 C10H16O m/z=152.1151 (+)-3-Thujone Met35762 HMDB36113 152.1201151 C10H16O m/z=152.1151 (-)-3-Thujone Met35761 HMDB36114 152.1201151 C10H16O m/z=152.1151 (-)-3-Isothujone Met35760 HMDB36115 152.1201151 C10H16O m/z=152.1151 (-)-Pinocamphone Met22313 HMDB36127 152.1201151 C10H16O m/z=152.1151 (-)-trans-Pinocarveol Met11120 HMDB36128 152.1201151 C10H16O m/z=152.1151 Verbenol Met11121 HMDB36129 152.1201151 C10H16O m/z=152.1151 (E,E)-2,4-Decadienal Met23649 HMDB36598 152.1201151 C10H16O m/z=152.1151 6,8-Epoxy-p-menth-2-ene Met17005 HMDB37008 152.1201151 C10H16O m/z=152.1151 Terpinolene oxide Met17006 HMDB37009 152.1201151 C10H16O m/z=152.1151 (2S,4R)-p-Mentha-1(7),8-dien-2-ol Met31479 HMDB37010 152.1201151 C10H16O m/z=152.1151 p-Mentha-1,8-dien-4-ol Met31482 HMDB37015 152.1201151 C10H16O m/z=152.1151 (R)-p-Mentha-1,8-dien-10-ol Met31484 HMDB37017 152.1201151 C10H16O bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

m/z=152.1151 4-Acetyl-1,4-dimethyl-1-cyclohexene Met27090 HMDB37024 152.1201151 C10H16O m/z=152.1151 Hop ether Met27091 HMDB37025 152.1201151 C10H16O Tetrahydro-5-isopropenyl-2-methyl-2- m/z=152.1151 Met21750 HMDB37172 152.1201151 C10H16O vinylfuran m/z=152.1151 Junionone Met14054 HMDB37280 152.1201151 C10H16O m/z=152.1151 xi-p-Mentha-3,8-dien-1-ol Met42772 HMDB37302 152.1201151 C10H16O m/z=152.1151 xi-p-Mentha-1(7),2-dien-4-ol Met42771 HMDB37303 152.1201151 C10H16O m/z=152.1151 xi-Pinol Met13963 HMDB38251 152.1201151 C10H16O m/z=152.1151 Photocitral A Met34260 HMDB38290 152.1201151 C10H16O xi-3,6-Dihydro-4-methyl-2-(2-methyl-1- m/z=152.1151 Met42277 HMDB38557 152.1201151 C10H16O propenyl)-2H-pyran m/z=152.1151 3,4-Epoxy-p-menth-1(7)-ene Met37711 HMDB39008 152.1201151 C10H16O m/z=152.1151 2-Hexylfuran Met25270 HMDB39816 152.1201151 C10H16O m/z=152.1151 Anethofuran Met28888 HMDB40766 152.1201151 C10H16O m/z=152.1151 beta-Cyclocitral Met43579 HMDB41011 152.1201151 C10H16O m/z=152.1151 Isocyclocitral Met43581 HMDB41013 152.1201151 C10H16O m/z=152.1151 3-Thujanone Met13671 HMDB41631 152.1201151 C10H16O m/z=153.1077 psi-Pelletierine Met12508 HMDB34580 153.1153641 C9H15NO m/z=153.1077 2-Methyl-4-pentyloxazole Met13683 HMDB37856 153.1153641 C9H15NO m/z=153.1077 4-Methyl-2-pentyloxazole Met13678 HMDB37857 153.1153641 C9H15NO m/z=153.1077 5-Butyl-2-ethyloxazole Met18237 HMDB37862 153.1153641 C9H15NO m/z=153.1077 2-Ethyl-4-methyl-5-propyloxazole Met32706 HMDB37872 153.1153641 C9H15NO m/z=153.1077 4-Ethyl-2-methyl-5-propyloxazole Met14624 HMDB37873 153.1153641 C9H15NO m/z=153.1077 5-Ethyl-4-methyl-2-(1-methylethyl)oxazole Met14641 HMDB37874 153.1153641 C9H15NO m/z=153.1077 2-Butyl-4,5-dimethyloxazole Met32710 HMDB37875 153.1153641 C9H15NO m/z=153.1077 5-Ethyl-2-methyl-4-propyloxazole Met32709 HMDB37876 153.1153641 C9H15NO m/z=153.1077 4-Ethyl-5-methyl-2-(1-methylethyl)oxazole Met32713 HMDB37878 153.1153641 C9H15NO m/z=153.1077 4-Butyl-2,5-dimethyloxazole Met32712 HMDB37879 153.1153641 C9H15NO m/z=153.1077 2-Isobutyl-4,5-dimethyloxazole Met25548 HMDB37880 153.1153641 C9H15NO m/z=153.1077 4-Ethyl-5-methyl-2-propyloxazole Met25549 HMDB37881 153.1153641 C9H15NO m/z=153.1077 5-Butyl-2,4-dimethyloxazole Met25546 HMDB37882 153.1153641 C9H15NO m/z=154.0058 1,4-Dithiothreitol Met915 HMDB13593 154.0122209 C4H10O2S2 m/z=154.0058 1-(2-Thienyl)-1,2-propanedione Met17124 HMDB40002 154.0088501 C7H6O2S m/z=155.1303 Methylisopelletierine Met35080 HMDB30326 155.1310142 C9H17NO m/z=155.1303 2,2,6,6-Tetramethyl-4-piperidinone Met6299 HMDB31179 155.1310142 C9H17NO m/z=155.1303 3-(1-Pyrrolidinyl)-2-pentanone Met17983 HMDB39833 155.1310142 C9H17NO m/z=155.1303 2-(1-Pyrrolidinyl)-3-pentanone Met17978 HMDB39834 155.1310142 C9H17NO m/z=156.1592 Menthol Met5210 HMDB03352 156.1514153 C10H20O m/z=156.1592 Decanal Met43897 HMDB11623 156.1514153 C10H20O m/z=156.1592 (E)-3-decen-1-ol Met44294 HMDB13810 156.1514153 C10H20O m/z=156.1592 2-Decanone Met39704 HMDB31409 156.1514153 C10H20O m/z=156.1592 cis-4-Decenol Met20669 HMDB32206 156.1514153 C10H20O bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

m/z=156.1592 3-Decanone Met6196 HMDB32212 156.1514153 C10H20O m/z=156.1592 3,7-Dimethyloctanal Met41158 HMDB32241 156.1514153 C10H20O m/z=156.1592 (+/-)-4-Ethyloctanal Met16353 HMDB32274 156.1514153 C10H20O m/z=156.1592 2-DECENOL Met26589 HMDB32532 156.1514153 C10H20O m/z=156.1592 Menthanol Met24156 HMDB34717 156.1514153 C10H20O m/z=156.1592 D-Citronellol Met44108 HMDB35093 156.1514153 C10H20O m/z=156.1592 L-Citronellol Met44107 HMDB35094 156.1514153 C10H20O m/z=156.1592 p-Menthan-4-ol Met16455 HMDB35726 156.1514153 C10H20O m/z=156.1592 (+)-Neomenthol Met36747 HMDB35763 156.1514153 C10H20O m/z=156.1592 (-)-Neoisomenthol Met36742 HMDB35764 156.1514153 C10H20O m/z=156.1592 p-Menthan-3-ol Met36741 HMDB35765 156.1514153 C10H20O m/z=156.1592 p-Menthan-1-ol Met27086 HMDB37020 156.1514153 C10H20O m/z=156.1592 alpha-Citronellol Met26733 HMDB37171 156.1514153 C10H20O m/z=156.1592 (±)-Carvomenthol Met5330 HMDB37223 156.1514153 C10H20O m/z=156.1592 Dihydrocitronellal Met19268 HMDB37806 156.1514153 C10H20O m/z=156.1592 1-Decen-3-ol Met7693 HMDB39854 156.1514153 C10H20O m/z=156.1592 (+)-Neoisomenthol Met20995 HMDB41628 156.1514153 C10H20O m/z=180.0193 2,4,6-Trimethyl-1,3,5-trithiane Met24561 HMDB31671 180.0101125 C6H12S3 m/z=180.0193 1-(Methylthio)ethyl 2-propenyl disulfide Met13906 HMDB32115 180.0101125 C6H12S3 m/z=180.0193 2,5-Dimethyl-1,4-dithiane-2,5-diol Met32769 HMDB33556 180.027871 C6H12O2S2 m/z=180.0193 Ethyl 2-(methyldithio)propionate Met28969 HMDB38896 180.027871 C6H12O2S2 m/z=184.0502 (+/-)-3-[(2-methyl-3-furyl)thio]-2-butanone Met10146 HMDB32401 184.0558003 C9H12O2S m/z=192.1189 2-Methyl-1-phenyl-2-propanyl acetate Met43497 HMDB31571 192.1150298 C12H16O2 m/z=192.1189 4-Phenyl-2-butyl acetate Met33335 HMDB31614 192.1150298 C12H16O2 m/z=192.1189 Ethyl 4-phenylbutanoate Met33327 HMDB31618 192.1150298 C12H16O2 m/z=192.1189 Benzyl 3-methylbutanoate Met38261 HMDB32043 192.1150298 C12H16O2 m/z=192.1189 3-Methylbutyl benzoate Met11957 HMDB33380 192.1150298 C12H16O2 m/z=192.1189 2-Methylpropyl phenylacetate Met42697 HMDB35010 192.1150298 C12H16O2 m/z=192.1189 2-Phenylethyl butanoate Met42693 HMDB35014 192.1150298 C12H16O2 m/z=192.1189 2-Phenylethyl 2-methylpropanoate Met42692 HMDB35015 192.1150298 C12H16O2 m/z=192.1189 5-Isopropyl-2-methylphenol acetate Met35852 HMDB35447 192.1150298 C12H16O2 m/z=192.1189 3-Phenylpropyl propanoate Met24154 HMDB36388 192.1150298 C12H16O2 11,12,13-Trinor-1(10)-spirovetivene-2,7- m/z=192.1189 Met12301 HMDB36734 192.1150298 C12H16O2 dione 4-Methyl-2-(1-phenylethyl)-1,3-dioxolane, m/z=192.1189 Met36329 HMDB37165 192.1150298 C12H16O2 9CI m/z=192.1189 4-Methylphenyl 3-methylbutanoate Met29043 HMDB37709 192.1150298 C12H16O2 m/z=192.1189 1-Phenylethyl isobutyrate Met42984 HMDB37716 192.1150298 C12H16O2 m/z=192.1189 1-Phenylethyl butyrate Met14511 HMDB37717 192.1150298 C12H16O2 1-Ethoxy-2-methoxy-4-(1- m/z=192.1189 Met37915 HMDB37792 192.1150298 C12H16O2 propenyl)benzene 2-(2-Hydroxy-4-methylphenyl)-3- m/z=192.1189 Met43220 HMDB38180 192.1150298 C12H16O2 pentanone bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

2-Ethoxy-1-methoxy-4-(1- m/z=192.1189 Met30335 HMDB39496 192.1150298 C12H16O2 propenyl)benzene 11,12,13-Trinor-1,3,5-bisabolatrien-10-oic m/z=192.1189 Met14609 HMDB39643 192.1150298 C12H16O2 acid m/z=192.1189 2-Benzyl-4,5-dimethyl-1,3-dioxolane Met32417 HMDB39841 192.1150298 C12H16O2 m/z=192.1189 Butyl phenylacetate Met37305 HMDB40427 192.1150298 C12H16O2 m/z=194.1152 Laccarin Met32640 HMDB41440 194.1055277 C10H14N2O2 m/z=196.0903 2-(Ethylamino)-4,5-dihydroxybenzamide Met16648 HMDB32852 196.0847923 C9H12N2O3 m/z=196.0903 5-Nitro-2-propoxyaniline Met42666 HMDB37688 196.0847923 C9H12N2O3 m/z=196.0903 1-(2-Thienyl)-1-heptanone Met37759 HMDB40241 196.0921858 C11H16OS m/z=200.0438 Potassium 2-(1'-ethoxy) ethoxypropanoate Met11645 HMDB37192 200.0450908 C7H13KO4 m/z=200.0438 Camalexin Met12815 HMDB38631 200.040819 C11H8N2S m/z=200.0443 Potassium 2-(1'-ethoxy) ethoxypropanoate Met11645 HMDB37192 200.0450908 C7H13KO4 m/z=200.0443 Camalexin Met12815 HMDB38631 200.040819 C11H8N2S m/z=200.046 Potassium 2-(1'-ethoxy) ethoxypropanoate Met11645 HMDB37192 200.0450908 C7H13KO4 m/z=200.046 Camalexin Met12815 HMDB38631 200.040819 C11H8N2S m/z=200.1675 2-Hexenoylcholine Met9828 HMDB13162 200.165054 C11H22NO2 m/z=218.0129 Diphenyl disulfide Met43013 HMDB31823 218.0223917 C12H10S2 m/z=236.0509 Aspartyl-Cysteine Met40029 HMDB28750 236.0466922 C7H12N2O5S m/z=236.0509 Cysteinyl-Aspartate Met10784 HMDB28771 236.0466922 C7H12N2O5S m/z=236.0509 Brassinin Met17767 HMDB33350 236.0441898 C11H12N2S2 m/z=242.022 Inositol cyclic phosphate Met38558 HMDB01125 242.0191538 C6H11O8P m/z=246.0089 Dimethylallylpyrophosphate Met2437 HMDB01120 246.0058258 C5H12O7P2 m/z=246.0089 Dimethylallylpyrophosphate Met2439 HMDB01120 246.0058258 C5H12O7P2 m/z=246.0089 Isopentenyl pyrophosphate Met2405 HMDB01347 246.0058258 C5H12O7P2 m/z=246.0089 Isopentenyl pyrophosphate Met2418 HMDB01347 246.0058258 C5H12O7P2 m/z=246.0089 Isopentenyl pyrophosphate Met2413 HMDB01347 246.0058258 C5H12O7P2 m/z=246.2208 8,8-Diethoxy-2,6-dimethyl-2-octanol Met6723 HMDB34557 246.2194948 C14H30O3 m/z=251.2239 Herculin Met9833 HMDB30275 251.2249146 C16H29NO m/z=254.1321 Homoanserine Met5820 HMDB05767 254.1378905 C11H18N4O3 m/z=254.1321 Histidinyl-Valine Met7272 HMDB28898 254.1378905 C11H18N4O3 m/z=254.1321 Valyl-Histidine Met32174 HMDB29129 254.1378905 C11H18N4O3 m/z=254.1321 L-Furosine Met9726 HMDB29390 254.1266571 C12H18N2O4 m/z=254.1321 L-Pyridosine Met20884 HMDB29443 254.1266571 C12H18N2O4 m/z=254.1321 Isoeugenol benzyl ether Met37709 HMDB32058 254.1306798 C17H18O2 m/z=254.1321 Isoeugenyl benzyl ether Met10391 HMDB32349 254.1306798 C17H18O2 m/z=254.1321 Pyrraline Met35129 HMDB33143 254.1266571 C12H18N2O4 m/z=254.1321 (+)-Setoclavine Met11401 HMDB33428 254.1419132 C16H18N2O m/z=254.1321 4-[(2-Methyl-3-furanyl)thio]-5-nonanone Met7189 HMDB37141 254.1340506 C14H22O2S 2,6-Dimethyl-3-[(2-methyl-3-furanyl)thio]- m/z=254.1321 Met31892 HMDB37154 254.1340506 C14H22O2S 4-heptanone m/z=254.1321 (S)-Argpyrimidine Met26147 HMDB37180 254.1378905 C11H18N4O3 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

m/z=254.1321 Aldosine Met28069 HMDB37817 254.1266571 C12H18N2O4 m/z=255.2543 Palmitic amide Met11268 HMDB12273 255.2562147 C16H33NO m/z=258.131 Strobilurin A Met43837 HMDB30420 258.1255944 C16H18O3 m/z=258.131 gamma-L-Glutamyl-L-pipecolic acid Met42048 HMDB38614 258.1215717 C11H18N2O5 m/z=258.131 (2S,2'S)-Pyrosaccharopine Met31799 HMDB38676 258.1215717 C11H18N2O5 6,10,14-Trimethyl-5,9,13-pentadecatrien- m/z=262.2268 Met9412 HMDB34495 262.2296656 C18H30O 2-one m/z=262.2268 2-(5,8-Tetradecadienyl)cyclobutanone Met38859 HMDB37519 262.2296656 C18H30O 5,8-Dihydro-6-(4-methyl-3-pentenyl)- m/z=264.0141 Met10211 HMDB38142 264.0134833 C10H16S4 1,2,3,4-tetrathiocin m/z=274.0626 cis,trans-5'-Hydroxythalidomide Met21705 HMDB13870 274.0589714 C13H10N2O5 m/z=274.0626 5-Hydroxythalidomide Met15962 HMDB13871 274.0589714 C13H10N2O5 m/z=274.0626 Thalidomide arene oxide Met15986 HMDB13872 274.0589714 C13H10N2O5 4-Phenyl-1H,3H-naphtho[1,8-cd]pyran-1,3- m/z=274.0626 Met8903 HMDB41120 274.0629942 C18H10O3 dione m/z=274.1871 3-Hydroxytetradecanedioic acid Met13609 HMDB00394 274.1780239 C14H26O5 m/z=274.1871 Nandrolone Met25883 HMDB02725 274.1932801 C18H26O2 m/z=274.1871 1-Phenyl-1,3-dodecanedione Met39714 HMDB35579 274.1932801 C18H26O2 (10Z,14E,16E)-10,14,16-Octadecatrien-12- m/z=274.1871 Met41952 HMDB35963 274.1932801 C18H26O2 ynoic acid m/z=274.1871 Rhodinyl phenylacetate Met11644 HMDB37191 274.1932801 C18H26O2 m/z=274.1871 Citronellyl alpha-toluate Met19793 HMDB37230 274.1932801 C18H26O2 m/z=281.2698 Oleamide Met39124 HMDB02117 281.2718648 C18H35NO m/z=283.173 Histidinyl-Lysine Met7275 HMDB28890 283.1644396 C12H21N5O3 m/z=283.173 Lysyl-Histidine Met25601 HMDB28953 283.1644396 C12H21N5O3 m/z=286.1186 Histidinyl-Methionine Met7274 HMDB28891 286.1099612 C11H18N4O3S m/z=286.1186 Methionyl-Histidine Met15558 HMDB28975 286.1099612 C11H18N4O3S m/z=286.1186 4'-Hydroxy-5,7-dimethoxyflavan Met20248 HMDB30152 286.1205091 C17H18O4 m/z=286.1186 Sativan Met44089 HMDB30521 286.1205091 C17H18O4 m/z=286.1186 Isosativan Met37353 HMDB34024 286.1205091 C17H18O4 m/z=286.1186 Myrigalone G Met32615 HMDB41168 286.1205091 C17H18O4 m/z=286.1186 Myrigalone H Met32616 HMDB41169 286.1205091 C17H18O4 m/z=286.1186 4'-Hydroxy-3,4,5-trimethoxystilbene Met6317 HMDB41671 286.1205091 C17H18O4 m/z=286.1186 4-Hydroxy-3,5,4'-trimethoxystilbene Met9446 HMDB41680 286.1205091 C17H18O4 m/z=297.2635 Palmitoleoyl Ethanolamide Met31023 HMDB13648 297.2667794 C18H35NO2 m/z=316.0932 Homoferreirin Met39413 HMDB30111 316.0946882 C17H16O6 3-(3,4-Dihydroxybenzyl)-7-hydroxy-5- m/z=316.0932 Met38763 HMDB33298 316.0946882 C17H16O6 methoxy-4-chromanone m/z=316.0932 Melilotocarpan E Met17808 HMDB33683 316.0946882 C17H16O6 m/z=316.0932 Melilotocarpan D Met32260 HMDB33692 316.0946882 C17H16O6 m/z=316.0932 Olivin Met30291 HMDB33797 316.0946882 C17H16O6 m/z=316.0932 Cajanol Met17304 HMDB33924 316.0946882 C17H16O6 5,8-Dihydroxy-3-(4-hydroxybenzyl)-7- m/z=316.0932 Met11087 HMDB37251 316.0946882 C17H16O6 methoxy-4-chromanone bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

5,7-Dihydroxy-3-(3-hydroxy-4- m/z=316.0932 Met30318 HMDB37477 316.0946882 C17H16O6 methoxybenzyl)-4-chromanone m/z=316.0932 3',5-Dihydroxy-4',7-dimethoxyflavanone Met24610 HMDB37481 316.0946882 C17H16O6 m/z=316.0932 4',7-Dihydroxy-2',5-dimethoxyisoflavanone Met8672 HMDB37749 316.0946882 C17H16O6 m/z=316.0932 Artocarpanone A Met36797 HMDB41234 316.0946882 C17H16O6 m/z=316.0932 Violanone Met34409 HMDB41790 316.0946882 C17H16O6 m/z=318.1042 4',7-Di-O-methylcatechin Met26607 HMDB30663 318.1103383 C17H18O6 m/z=323.2731 Linoleoyl ethanolamide Met40529 HMDB12252 323.2824294 C20H37NO2 m/z=333.2133 Mahanimbinol Met21128 HMDB40992 333.2092645 C23H27NO m/z=336.0738 S-Nitrosoglutathione Met25835 HMDB04645 336.0739696 C10H16N4O7S m/z=344.0792 Thiamine monophosphate Met1902 HMDB02666 344.0708123 C12H17N4O4PS m/z=344.0792 Thiamine monophosphate Met2957 HMDB02666 344.0708123 C12H17N4O4PS m/z=344.0792 Thiamine monophosphate Met1365 HMDB02666 344.0708123 C12H17N4O4PS m/z=344.0792 Theogallin Met43712 HMDB39287 344.0743467 C14H16O10 N-[2-(3,4-Dimethoxyphenyl)ethyl]-3,4- m/z=371.174 Met10634 HMDB32226 371.1732729 C21H25NO5 dimethoxycinnamic acid amide m/z=374.0725 Neodiospyrin Met18499 HMDB29538 374.0790382 C22H14O6 m/z=382.0746 (+)-12a-Hydroxypachyrrhizone Met32819 HMDB30774 382.0688674 C20H14O8 m/z=385.2813 3-Hydroxy-cis-5-tetradecenoylcarnitine Met28143 HMDB13330 385.2828234 C21H39NO5 m/z=386.1856 Corchoionol C 9-glucoside Met42740 HMDB29772 386.1940679 C19H30O8 m/z=386.1856 Citroside A Met30704 HMDB30370 386.1940679 C19H30O8 m/z=386.1856 Sonchuionoside C Met24981 HMDB35212 386.1940679 C19H30O8 m/z=386.1857 Corchoionol C 9-glucoside Met42740 HMDB29772 386.1940679 C19H30O8 m/z=386.1857 Citroside A Met30704 HMDB30370 386.1940679 C19H30O8 m/z=386.1857 Sonchuionoside C Met24981 HMDB35212 386.1940679 C19H30O8 m/z=386.9697 5'-Hydroxylornoxicam Met8177 HMDB13991 386.9750395 C13H10ClN3O5S2 m/z=387.356 25-Azacholesterol Met23956 HMDB01028 387.3501151 C26H45NO m/z=388.0988 Dopaxanthin quinone Met6854 HMDB12220 388.0906655 C18H16N2O8 m/z=388.0988 Betanidin Met25803 HMDB29407 388.0906655 C18H16N2O8 m/z=404.1957 Pisumionoside Met20330 HMDB39947 404.2046326 C19H32O9 m/z=422.1628 3-(1,1-Dimethylallyl)scopoletin 7-glucoside Met13486 HMDB32853 422.1576824 C21H26O9 m/z=422.2871 DHAP(18:0e) Met19891 HMDB11142 422.2797256 C21H43O6P m/z=424.3357 Alpha-Tocotrienol Met1076 HMDB06327 424.3341307 C29H44O2 m/z=424.3357 Alpha-Tocotrienol Met2558 HMDB06327 424.3341307 C29H44O2 m/z=424.3357 Alpha-Tocotrienol Met2552 HMDB06327 424.3341307 C29H44O2 m/z=424.3357 (22E,24R)-Stigmasta-4,22-diene-3,6-dione Met21656 HMDB38656 424.3341307 C29H44O2 m/z=426.1673 Mangostanol Met18559 HMDB29868 426.1678532 C24H26O7 m/z=426.1673 Edulisin II Met7596 HMDB30593 426.1678532 C24H26O7 m/z=426.1673 Archangelicin Met41721 HMDB30845 426.1678532 C24H26O7 m/z=426.1673 Atrovirisidone Met10067 HMDB36982 426.1678532 C24H26O7 m/z=426.1673 Mangostenol Met26140 HMDB39912 426.1678532 C24H26O7 m/z=431.1047 Ochratoxin C Met35768 HMDB29400 431.1135652 C22H22ClNO6 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

m/z=432.1976 S-Furanopetasitin Met25564 HMDB36131 432.1970448 C24H32O5S m/z=434.0375 Halosulfuron-methyl Met18471 HMDB34859 434.0411453 C13H15ClN6O7S m/z=434.1183 2-O-Caffeoylarbutin Met36160 HMDB31347 434.1212969 C21H22O10 m/z=434.1183 Floribundoside Met39355 HMDB33739 434.1212969 C21H22O10 m/z=434.1183 Eriodictin Met24609 HMDB37480 434.1212969 C21H22O10 m/z=434.1183 5,7,8-Trihydroxyflavanone 7-glucoside Met13397 HMDB38775 434.1212969 C21H22O10 m/z=434.1183 Dihydrogenistin Met36318 HMDB39935 434.1212969 C21H22O10 (2R,3R)-3,3',4',7-Tetrahydroxyflavanone 7- m/z=434.1183 Met30878 HMDB40536 434.1212969 C21H22O10 O-alpha-L-Rhamnopyranoside m/z=434.1183 5-Hydroxyaloin A Met43889 HMDB40832 434.1212969 C21H22O10 m/z=434.1183 7-Hydroxyaloin B Met33602 HMDB41150 434.1212969 C21H22O10 7-Hydroxy-5-(4-hydroxy-2-oxopentyl)-2- m/z=438.1474 Met20842 HMDB34698 438.1525971 C21H26O10 methylchromone 7-glucoside m/z=438.1474 Cnidimol 7-glucoside Met35258 HMDB40502 438.1525971 C21H26O10 m/z=448.1435 Isosakuranin Met42797 HMDB29481 448.136947 C22H24O10 m/z=448.1435 Cycloartomunoxanthone Met30971 HMDB29999 448.1522031 C26H24O7 m/z=448.1435 Artonol E Met5688 HMDB30490 448.1522031 C26H24O7 m/z=448.1435 7-Hydroxy-8-O-methylaloin B Met26013 HMDB31586 448.136947 C22H24O10 m/z=448.1435 Puddumin A Met44275 HMDB33741 448.136947 C22H24O10 m/z=448.1435 Chalconosakuranetin Met13950 HMDB37509 448.136947 C22H24O10 m/z=448.1435 Cycloartomunin Met39344 HMDB38878 448.1522031 C26H24O7 Aromadendrin 4'-methyl ether 7- m/z=448.1435 Met26471 HMDB40561 448.136947 C22H24O10 rhamnoside m/z=448.1435 Piperenol C Met9896 HMDB41536 448.136947 C22H24O10 m/z=450.3381 Coprocholic acid Met9074 HMDB00601 450.3345246 C27H46O5 m/z=450.3381 3a,7a,12a-Trihydroxy-5b-cholestanoic acid Met4701 HMDB03873 450.3345246 C27H46O5 m/z=450.3381 3a,7a,12a-Trihydroxy-5b-cholestanoic acid Met2786 HMDB03873 450.3345246 C27H46O5 m/z=450.3381 3a,7a,12a-Trihydroxy-5b-cholestanoic acid Met2871 HMDB03873 450.3345246 C27H46O5 m/z=450.3381 3a,7a,12a-Trihydroxy-5b-cholestanoic acid Met2840 HMDB03873 450.3345246 C27H46O5 m/z=463.0728 Adenylsuccinic acid Met4101 HMDB00536 463.074043 C14H18N5O11P m/z=463.0728 Chondroitin sulfate Met10372 HMDB00580 463.0631897 C13H21NO15S m/z=464.1005 Quercetin 3-galactoside Met32818 HMDB30775 464.0954761 C21H20O12 m/z=464.1005 Myricetin 7-rhamnoside Met29412 HMDB33132 464.0954761 C21H20O12 m/z=464.1005 Myricitrin Met44352 HMDB34360 464.0954761 C21H20O12 m/z=464.1005 6-Hydroxykaempferol 7-glucoside Met38941 HMDB34725 464.0954761 C21H20O12 m/z=464.1005 2'-Hydroxyisoorientin Met22090 HMDB35883 464.0954761 C21H20O12 m/z=464.1005 Isoquercitrin Met33749 HMDB37362 464.0954761 C21H20O12 m/z=464.1005 Quercetin 4'-glucoside Met29446 HMDB37932 464.0954761 C21H20O12 m/z=464.1005 6-Hydroxykaempferol 3-glucoside Met32727 HMDB40478 464.0954761 C21H20O12 m/z=465.1699 (E)-Squamosamide Met21167 HMDB41088 465.1787522 C26H27NO7 m/z=467.1838 Bis-N-butyl phthalate Met20289 HMDB31949 467.1791462 C22H29NO10 m/z=468.1693 Dukunolide D Met41260 HMDB35732 468.1784179 C26H28O8 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

m/z=468.1693 Jangomolide Met40272 HMDB38158 468.1784179 C26H28O8 m/z=472.28 Ixocarpanolide Met13411 HMDB33899 472.282489 C28H40O6 m/z=472.28 Vamonolide Met19290 HMDB37379 472.282489 C28H40O6 m/z=472.28 Macrocarpal B Met33078 HMDB38704 472.282489 C28H40O6 m/z=472.28 Macrocarpal D Met33653 HMDB38733 472.282489 C28H40O6 m/z=472.28 Macrocarpal E Met33648 HMDB38734 472.282489 C28H40O6 m/z=472.28 Macrocarpal H Met25733 HMDB41586 472.282489 C28H40O6 m/z=476.2152 Loperamide Met11784 HMDB04999 476.223056 C29H33ClN2O2 m/z=478.0686 Quercetin 3-O-glucuronide Met7973 HMDB29212 478.0747407 C21H18O13 m/z=478.0686 Quercetin-4'-glucuronide Met7967 HMDB29214 478.0747407 C21H18O13 m/z=478.0686 Quercetin 4'-glucuronide Met16338 HMDB33668 478.0747407 C21H18O13 m/z=478.0686 Neoglucobrassicin Met21681 HMDB38384 478.0715863 C17H22N2O10S2 m/z=478.0686 Quercetin 3'-O-glucuronide Met31693 HMDB41769 478.0747407 C21H18O13 m/z=478.0729 Quercetin 3-O-glucuronide Met7973 HMDB29212 478.0747407 C21H18O13 m/z=478.0729 Quercetin-4'-glucuronide Met7967 HMDB29214 478.0747407 C21H18O13 m/z=478.0729 Quercetin 4'-glucuronide Met16338 HMDB33668 478.0747407 C21H18O13 m/z=478.0729 Neoglucobrassicin Met21681 HMDB38384 478.0715863 C17H22N2O10S2 m/z=478.0729 Quercetin 3'-O-glucuronide Met31693 HMDB41769 478.0747407 C21H18O13 m/z=482.1805 Melleolide D Met37428 HMDB37042 482.1707457 C24H31ClO8 m/z=485.2783 Sarcodon scabrosus Depsipeptide Met41665 HMDB41421 485.2737152 C23H39N3O8 m/z=490.2027 Cymorcin diglucoside Met31409 HMDB39386 490.2050266 C22H34O12 m/z=492.1346 Tricin 7-glucoside Met27923 HMDB30553 492.1267762 C23H24O12 m/z=492.1346 Tricin 7-glucoside Met27923 HMDB30553 492.1267762 C23H24O12 m/z=492.1346 Licoagroside A Met16811 HMDB35448 492.1267762 C23H24O12 m/z=492.1346 Licoagroside A Met16811 HMDB35448 492.1267762 C23H24O12 m/z=492.1346 Americanin B Met38443 HMDB37338 492.1420324 C27H24O9 m/z=492.1346 Americanin B Met38443 HMDB37338 492.1420324 C27H24O9 m/z=492.1346 Limocitrin 3-rhamnoside Met9534 HMDB37578 492.1267762 C23H24O12 m/z=492.1346 Limocitrin 3-rhamnoside Met9534 HMDB37578 492.1267762 C23H24O12 m/z=492.1346 3',8-Dimethoxyapigenin 7-glucoside Met26361 HMDB38825 492.1267762 C23H24O12 m/z=492.1346 3',8-Dimethoxyapigenin 7-glucoside Met26361 HMDB38825 492.1267762 C23H24O12 3',4',5-Trihydroxy-3,7-dimethoxyflavone 5- m/z=492.1346 Met5398 HMDB39336 492.1267762 C23H24O12 glucoside 3',4',5-Trihydroxy-3,7-dimethoxyflavone 5- m/z=492.1346 Met5398 HMDB39336 492.1267762 C23H24O12 glucoside 3',7-Dimethoxy-4',5,8-trihydroxyflavone 8- m/z=492.1346 Met17184 HMDB40479 492.1267762 C23H24O12 glucoside 3',7-Dimethoxy-4',5,8-trihydroxyflavone 8- m/z=492.1346 Met17184 HMDB40479 492.1267762 C23H24O12 glucoside m/z=500.174 Artonol C Met20211 HMDB30489 500.1835033 C30H28O7 m/z=504.1671 Maltotriose Met3968 HMDB01262 504.169035 C18H32O16 m/z=504.1671 Maltotriose Met3970 HMDB01262 504.169035 C18H32O16 m/z=504.1671 Maltotriose Met3976 HMDB01262 504.169035 C18H32O16 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

m/z=504.1671 Levan Met38271 HMDB03539 504.169035 C18H32O16 m/z=504.1671 3-Galactosyllactose Met41497 HMDB06599 504.169035 C18H32O16 m/z=504.1671 Dextrin Met20460 HMDB06857 504.169035 C18H32O16 m/z=504.1671 1-Kestose Met12650 HMDB11729 504.169035 C18H32O16 m/z=504.1671 Melezitose Met37559 HMDB11730 504.169035 C18H32O16 m/z=504.1671 Gentiotriose Met29594 HMDB29910 504.169035 C18H32O16 m/z=504.1671 Neokestose Met44400 HMDB29921 504.169035 C18H32O16 m/z=504.1671 Galactotriose Met44397 HMDB29922 504.169035 C18H32O16 m/z=504.1671 6-O-Glucosylmaltose Met44396 HMDB29929 504.169035 C18H32O16 beta-D-Fructofuranosyl alpha-D- m/z=504.1671 Met19430 HMDB29930 504.169035 C18H32O16 glucopyranosyl-(1->4)-D-glucopyranoside m/z=504.1671 Panose Met19435 HMDB29937 504.169035 C18H32O16 m/z=504.1671 6-Kestose Met41285 HMDB33673 504.169035 C18H32O16 m/z=504.1671 Umbelliferose Met44287 HMDB33748 504.169035 C18H32O16 m/z=504.1671 Gentianose Met18122 HMDB34072 504.169035 C18H32O16 m/z=504.1671 Fagopyritol B2 Met29877 HMDB35322 504.169035 C18H32O16 beta-D-Galactopyranosyl-(1->2)-[beta-D- m/z=504.1671 Met10057 HMDB38851 504.169035 C18H32O16 galactopyranosyl-(1->4)]-D-galactose beta-D-Galactopyranosyl-(1->4)-beta-D- m/z=504.1671 Met10058 HMDB38852 504.169035 C18H32O16 galactopyranosyl-(1->4)-D-galactose beta-D-Galactopyranosyl-(1->3)-beta-D- m/z=504.1671 Met10059 HMDB38853 504.169035 C18H32O16 galactopyranosyl-(1->6)-D-galactose m/z=504.1671 3-beta-Gentiobiosylglucose Met10422 HMDB39703 504.169035 C18H32O16 alpha-D-Glucopyranosyl-(1->6)-alpha-D- m/z=504.1671 Met16646 HMDB39704 504.169035 C18H32O16 glucopyranosyl-(1->2)-D-glucose m/z=504.1671 4-beta-Laminaribiosylglucose Met26591 HMDB39706 504.169035 C18H32O16 alpha-D-Glucopyranosyl-(1->4)-alpha-D- m/z=504.1671 Met16643 HMDB39707 504.169035 C18H32O16 glucopyranosyl-(1->6)-D-glucose m/z=504.1671 3-beta-Cellobiosylglucose Met26585 HMDB39708 504.169035 C18H32O16 m/z=504.1671 Fagopyritol A2 Met36236 HMDB40181 504.169035 C18H32O16 m/z=504.1671 Sophorotriose Met35677 HMDB40984 504.169035 C18H32O16 m/z=504.1671 Nephritogenoside Met9029 HMDB41948 504.169035 C18H32O16 m/z=508.2302 Acetyl-T2 Toxin Met24960 HMDB33164 508.2308474 C26H36O10 m/z=508.2302 Mozambioside Met37643 HMDB37002 508.2308474 C26H36O10 m/z=508.2302 Cafamarine Met17008 HMDB37003 508.2308474 C26H36O10 m/z=508.2302 Gibberellin A37 glucosyl ester Met42053 HMDB38611 508.2308474 C26H36O10 m/z=508.493 4-Hydroxy-16,18-tritriacontanedione Met37390 HMDB39536 508.4855459 C33H64O3 m/z=510.2325 Deuteroporphyrin IX Met26234 HMDB00579 510.2267055 C30H30N4O4 m/z=515.3007 Taurocholic acid Met3797 HMDB00036 515.2916735 C26H45NO7S m/z=515.3007 Taurocholic acid Met3788 HMDB00036 515.2916735 C26H45NO7S m/z=515.3007 Taurocholic acid Met3791 HMDB00036 515.2916735 C26H45NO7S m/z=515.3007 Tauroursocholic acid Met8768 HMDB00889 515.2916735 C26H45NO7S m/z=515.3007 Taurallocholic acid Met8253 HMDB00922 515.2916735 C26H45NO7S m/z=515.3007 Tauro-b-muricholic acid Met32974 HMDB00932 515.2916735 C26H45NO7S bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

m/z=515.3007 LysoPC(18:4(6Z,9Z,12Z,15Z)) Met37045 HMDB10389 515.3011893 C26H46NO7P m/z=515.3007 Taurohyocholate Met19012 HMDB11637 515.2916735 C26H45NO7S m/z=530.1977 Artonin Q Met8907 HMDB41124 530.1940679 C31H30O8 m/z=534.11 6''-Malonylastragalin Met43846 HMDB37434 534.1009554 C24H22O14 4',5-Dihydroxy-3,3'-dimethoxy-6,7- m/z=534.11 Met18558 HMDB37758 534.1009554 C24H22O14 methylenedioxyflavone 4'-glucuronide 3,5-Dihydroxy-3',4'-dimethoxy-6,7- m/z=534.11 Met38921 HMDB38266 534.1009554 C24H22O14 methylenedioxyflavone 3-glucuronide 2'-Hydroxygenistein 7-(6''- m/z=534.11 Met21854 HMDB39247 534.1009554 C24H22O14 malonylglucoside) m/z=534.5089 Reticulatamol Met30437 HMDB41023 534.501196 C35H66O3 m/z=556.1058 Ephedrannin A Met35363 HMDB37654 556.1005615 C30H20O11 m/z=574.2096 Tetrahydrofolyl-[Glu](2) Met36632 HMDB06825 574.2135746 C24H30N8O9 m/z=574.3184 Ganoderic acid alpha Met37182 HMDB33024 574.3141831 C32H46O9 m/z=574.3184 Ganoderic acid K Met10027 HMDB35986 574.3141831 C32H46O9 m/z=578.2066 8-Acetoxypinoresinol 4-glucoside Met24024 HMDB33283 578.1999412 C28H34O13 m/z=58.0419 Acetone Met4346 HMDB01659 58.04186481 C3H6O m/z=58.0419 Acetone Met4338 HMDB01659 58.04186481 C3H6O m/z=58.0419 Acetone Met4348 HMDB01659 58.04186481 C3H6O m/z=58.0419 Propanal Met40247 HMDB03366 58.04186481 C3H6O m/z=58.0419 Methyloxirane Met33100 HMDB31558 58.04186481 C3H6O m/z=58.0419 Allyl alcohol Met14489 HMDB31652 58.04186481 C3H6O m/z=580.2166 (+)-7-epi-Syringaresinol 4'-glucoside Met38918 HMDB38261 580.2155912 C28H36O13 m/z=586.2005 Physalin E acetate Met40824 HMDB31878 586.2050266 C30H34O12 m/z=596.179 Eriocitrin Met33985 HMDB05811 596.1741204 C27H32O15 m/z=596.179 Isorubrofusarin 10-gentiobioside Met36914 HMDB34609 596.1741204 C27H32O15 (2R)-6,8-Diglucopyranosyl-4',5,7- m/z=596.179 Met25947 HMDB37407 596.1741204 C27H32O15 trihydroxyflavanone m/z=596.179 Pinobanksin 5-[galactosyl-(1->4)-glucoside] Met19293 HMDB37506 596.1741204 C27H32O15 m/z=596.179 Rubrofusarin 6-gentiobioside Met31459 HMDB38478 596.1741204 C27H32O15 m/z=596.179 Cassiaside C Met42272 HMDB38550 596.1741204 C27H32O15 m/z=610.1973 Hesperetin 7-neohesperidoside Met28509 HMDB30748 610.1897704 C28H34O15 m/z=610.1973 Protochlorophyllide Met10688 HMDB31148 610.206662 C35H30MgN4O5 4'-Hydroxyacetophenone 4'-[4-hydroxy- m/z=610.1973 3,5-dimethoxybenzoyl-(->5)-apiosyl-(1->2)- Met24253 HMDB36333 610.1897704 C28H34O15 glucoside] m/z=612.1747 Safflomin A Met27930 HMDB30558 612.169035 C27H32O16 m/z=612.1747 Carthamidin 6,7-diglucoside Met15240 HMDB40125 612.169035 C27H32O16 m/z=612.1747 Aromadendrin 3,7-diglucoside Met24797 HMDB40559 612.169035 C27H32O16 m/z=612.1747 Hydroxysafflor yellow A Met38149 HMDB40677 612.169035 C27H32O16 m/z=612.4774 DG(14:0/22:6(4Z,7Z,10Z,13Z,16Z,19Z)/0:0) Met40929 HMDB07034 612.4753752 C39H64O5 m/z=612.4774 DG(14:1(9Z)/22:5(4Z,7Z,10Z,13Z,16Z)/0:0) Met34936 HMDB07061 612.4753752 C39H64O5 m/z=612.4774 DG(14:1(9Z)/22:5(7Z,10Z,13Z,16Z,19Z)/0:0) Met34935 HMDB07062 612.4753752 C39H64O5 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

m/z=612.4774 DG(16:1(9Z)/20:5(5Z,8Z,11Z,14Z,17Z)/0:0) Met37010 HMDB07143 612.4753752 C39H64O5 m/z=612.4774 DG(18:2(9Z,12Z)/18:4(6Z,9Z,12Z,15Z)/0:0) Met13130 HMDB07251 612.4753752 C39H64O5 m/z=612.4774 DG(18:3(6Z,9Z,12Z)/18:3(6Z,9Z,12Z)/0:0) Met42408 HMDB07278 612.4753752 C39H64O5 m/z=612.4774 DG(18:3(6Z,9Z,12Z)/18:3(9Z,12Z,15Z)/0:0) Met42409 HMDB07279 612.4753752 C39H64O5 m/z=612.4774 DG(18:3(9Z,12Z,15Z)/18:3(6Z,9Z,12Z)/0:0) Met19464 HMDB07307 612.4753752 C39H64O5 m/z=612.4774 DG(18:3(9Z,12Z,15Z)/18:3(9Z,12Z,15Z)/0:0) Met13530 HMDB07308 612.4753752 C39H64O5 m/z=612.4774 DG(18:4(6Z,9Z,12Z,15Z)/18:2(9Z,12Z)/0:0) Met44444 HMDB07335 612.4753752 C39H64O5 m/z=612.4774 DG(20:5(5Z,8Z,11Z,14Z,17Z)/16:1(9Z)/0:0) Met10973 HMDB07563 612.4753752 C39H64O5 m/z=612.4774 DG(22:5(4Z,7Z,10Z,13Z,16Z)/14:1(9Z)/0:0) Met9665 HMDB07705 612.4753752 C39H64O5 m/z=612.4774 DG(22:5(7Z,10Z,13Z,16Z,19Z)/14:1(9Z)/0:0) Met18581 HMDB07734 612.4753752 C39H64O5 m/z=612.4774 DG(22:6(4Z,7Z,10Z,13Z,16Z,19Z)/14:0/0:0) Met24239 HMDB07762 612.4753752 C39H64O5 m/z=612.4774 DG(14:0/0:0/22:6n3) Met16232 HMDB55980 612.4753752 C39H64O5 m/z=612.4774 DG(14:1n5/0:0/22:5n6) Met33686 HMDB56149 612.4753752 C39H64O5 m/z=612.4774 DG(14:1n5/0:0/22:5n3) Met8999 HMDB56154 612.4753752 C39H64O5 m/z=612.4774 DG(16:1n7/0:0/20:5n3) Met19218 HMDB56174 612.4753752 C39H64O5 m/z=612.4774 DG(18:3n6/0:0/18:3n6) Met7353 HMDB56296 612.4753752 C39H64O5 m/z=612.4774 DG(18:3n6/0:0/18:3n3) Met30539 HMDB56303 612.4753752 C39H64O5 m/z=612.4774 DG(18:3n3/0:0/18:3n3) Met44458 HMDB56366 612.4753752 C39H64O5 m/z=614.1806 Catechin 3',4'-diglucoside Met19315 HMDB37950 614.184685 C27H34O16 m/z=614.1806 Catechin 3',5-diglucoside Met19316 HMDB37951 614.184685 C27H34O16 m/z=614.1806 Catechin 3',7-diglucoside Met19317 HMDB37952 614.184685 C27H34O16 m/z=614.1815 Catechin 3',4'-diglucoside Met19315 HMDB37950 614.184685 C27H34O16 m/z=614.1815 Catechin 3',5-diglucoside Met19316 HMDB37951 614.184685 C27H34O16 m/z=614.1815 Catechin 3',7-diglucoside Met19317 HMDB37952 614.184685 C27H34O16 m/z=629.3715 Janthitrem G Met19185 HMDB30531 629.3716384 C39H51NO6 m/z=633.1855 Pyranocyanin A Met26852 HMDB35420 633.1819454 C30H33O15 Quercetin 3-(4''-acetylrhamnoside) 7- m/z=636.1647 Met5395 HMDB39333 636.169035 C29H32O16 rhamnoside Kaempferol 3-(6''-acetylgalactoside) 7- m/z=636.1647 Met30881 HMDB40539 636.169035 C29H32O16 rhamnoside m/z=640.1353 Nelumboside Met6734 HMDB38464 640.1275641 C27H28O18 m/z=665.6623 Cer(d18:0/25:0) Met17087 HMDB11770 665.6685957 C43H87NO3 m/z=666.2185 Glycogen Met19005 HMDB00757 666.2218584 C24H42O21 m/z=666.2185 Maltotetraose Met621 HMDB01296 666.2218584 C24H42O21 m/z=666.2185 Maltotetraose Met1772 HMDB01296 666.2218584 C24H42O21 m/z=666.2185 Stachyose Met24619 HMDB03553 666.2218584 C24H42O21 m/z=666.2185 Mannan Met19429 HMDB29931 666.2218584 C24H42O21 m/z=666.2185 Fagopyritol B3 Met29876 HMDB35323 666.2218584 C24H42O21 m/z=666.2185 Bifurcose Met22746 HMDB38490 666.2218584 C24H42O21 m/z=666.2185 Neobifurcose Met39338 HMDB38870 666.2218584 C24H42O21 m/z=666.2185 Sesamose Met39339 HMDB38871 666.2218584 C24H42O21 m/z=666.2185 3-beta-Glucosylcellotriose Met13095 HMDB39173 666.2218584 C24H42O21 bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

m/z=666.2185 3-beta-Cellobiosylcellobiose Met25930 HMDB39174 666.2218584 C24H42O21 m/z=666.2185 Nystose Met25932 HMDB39176 666.2218584 C24H42O21 m/z=666.2185 Tetramethylquercetin 3-rutinoside Met5399 HMDB39337 666.2159852 C31H38O16 m/z=666.2185 Fagopyritol A3 Met36240 HMDB40182 666.2218584 C24H42O21 m/z=681.4365 PE(14:1(9Z)/18:4(6Z,9Z,12Z,15Z)) Met37225 HMDB08865 681.4369545 C37H64NO8P m/z=681.4365 PE(18:4(6Z,9Z,12Z,15Z)/14:1(9Z)) Met22585 HMDB09185 681.4369545 C37H64NO8P m/z=681.4365 7-Epizucchini factor A Met11227 HMDB36405 681.439324 C44H59NO5 m/z=683.4492 PE(14:0/18:4(6Z,9Z,12Z,15Z)) Met31322 HMDB08832 683.4526046 C37H66NO8P m/z=683.4492 PE(14:1(9Z)/18:3(6Z,9Z,12Z)) Met37227 HMDB08863 683.4526046 C37H66NO8P m/z=683.4492 PE(14:1(9Z)/18:3(9Z,12Z,15Z)) Met37224 HMDB08864 683.4526046 C37H66NO8P m/z=683.4492 PE(18:3(6Z,9Z,12Z)/14:1(9Z)) Met35726 HMDB09119 683.4526046 C37H66NO8P m/z=683.4492 PE(18:3(9Z,12Z,15Z)/14:1(9Z)) Met15454 HMDB09152 683.4526046 C37H66NO8P m/z=683.4492 PE(18:4(6Z,9Z,12Z,15Z)/14:0) Met22584 HMDB09184 683.4526046 C37H66NO8P m/z=684.2369 Citbismine C Met10175 HMDB41435 684.23191 C37H36N2O11 m/z=700.1371 Theaflavate B Met12352 HMDB33038 700.1428202 C36H28O15 m/z=700.1371 (S)-Skyrin 2-glucoside Met32886 HMDB36320 700.1428202 C36H28O15 m/z=717.5649 PC(14:0/P-18:0) Met21935 HMDB07896 717.5672404 C40H80NO7P m/z=717.5649 PC(16:0/P-16:0) Met23883 HMDB07994 717.5672404 C40H80NO7P m/z=717.5649 PC(P-16:0/16:0) Met28199 HMDB11206 717.5672404 C40H80NO7P m/z=717.5649 PC(P-18:0/14:0) Met32547 HMDB11236 717.5672404 C40H80NO7P m/z=717.5649 PC(o-16:0/16:1(9Z)) Met10609 HMDB13404 717.5672404 C40H80NO7P m/z=727.4437 PS(14:1(9Z)/18:3(9Z,12Z,15Z)) Met27683 HMDB12348 727.4424339 C38H66NO10P m/z=727.4437 PS(18:3(9Z,12Z,15Z)/14:1(9Z)) Met34377 HMDB12408 727.4424339 C38H66NO10P m/z=742.2466 Hexacarboxylporphyrin I Met33468 HMDB00743 742.2486227 C38H38N4O12 m/z=742.2466 Porphinehexacarboxylic acid Met14667 HMDB00780 742.2486227 C38H38N4O12 m/z=742.2466 Hexacarboxylporphyrin III Met5707 HMDB01952 742.2486227 C38H38N4O12 m/z=742.2469 Hexacarboxylporphyrin I Met33468 HMDB00743 742.2486227 C38H38N4O12 m/z=742.2469 Porphinehexacarboxylic acid Met14667 HMDB00780 742.2486227 C38H38N4O12 m/z=742.2469 Hexacarboxylporphyrin III Met5707 HMDB01952 742.2486227 C38H38N4O12 m/z=742.4758 PG(16:1(9Z)/18:3(6Z,9Z,12Z)) Met36410 HMDB10591 742.478485 C40H71O10P m/z=742.4758 PG(16:1(9Z)/18:3(9Z,12Z,15Z)) Met36409 HMDB10592 742.478485 C40H71O10P m/z=742.4758 PG(18:3(6Z,9Z,12Z)/16:1(9Z)) Met42825 HMDB10661 742.478485 C40H71O10P m/z=742.4758 PG(18:3(9Z,12Z,15Z)/16:1(9Z)) Met28150 HMDB10676 742.478485 C40H71O10P m/z=773.2238 Cyanidin 3,3',5-triglucoside Met9106 HMDB37973 773.2140334 C33H41O21 m/z=773.2238 Cyanidin 3-sophorotrioside Met41208 HMDB37995 773.2140334 C33H41O21 m/z=773.2238 Cyanidin 3-sophoroside 5-glucoside Met37398 HMDB38487 773.2140334 C33H41O21 m/z=773.2238 Cyanidin 3-diglucoside 5-glucoside Met37622 HMDB41716 773.2140334 C33H41O21 m/z=789.6309 PC(14:0/22:0) Met36513 HMDB07886 789.6247553 C44H88NO8P m/z=789.6309 PC(16:0/20:0) Met5403 HMDB07977 789.6247553 C44H88NO8P m/z=789.6309 PC(18:0/18:0) Met44165 HMDB08036 789.6247553 C44H88NO8P m/z=789.6309 PC(20:0/16:0) Met10185 HMDB08265 789.6247553 C44H88NO8P bioRxiv preprint doi: https://doi.org/10.1101/482422; this version posted December 1, 2018. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-ND 4.0 International license.

m/z=789.6309 PC(22:0/14:0) Met41375 HMDB08525 789.6247553 C44H88NO8P m/z=789.6309 PE(15:0/24:0) Met39815 HMDB08914 789.6247553 C44H88NO8P m/z=789.6309 PE(24:0/15:0) Met8004 HMDB09714 789.6247553 C44H88NO8P